The vRealize Operations Management Pack for Horizon collects metrics for objects within its plug-ins.
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Horizon Adapter Instance | Last Collection Time | collectorTime | Time at which the last collection happened for all collectors | |
| Last Event Time | eventEndTime | Time at which the last collection happened for events | ||
| Performance Statistics | Collection Duration | totalExecutionTime | Time taken to complete the Collection of selected Adapter Instance of VMware management pack for Horizon | |
| Updated Relationships | countOfUpdatedRelationships | Number of updated Relationships per collection cycle | ||
| Number of Events | countOfEvents | Total Events collected from Event DB | ||
| Data Collectors | Execution Time | Metric group for performance Statistics of Data Collectors of VMware management pack for Horizon | ||
| Performance Statistics/Data Collectors | Application Pool Resource Collector | applicationPoolResourceCollector | Time taken to complete the collection of Horizon Application Pools | |
| Application Pool Details Resource Collector | applicationPoolDetailsResourceCollector | Time taken to complete the collection of Horizon Application Pool Details | ||
| Application Session Details Resource Collector | applicationSessionDetailsResourceCollector | Time taken to complete the collection of Horizon Application Session Details | ||
| Application Session Resource Collector | applicationSessionResourceCollector | Time taken to complete the collection of Horizon Application Sessions | ||
| Connection Server Health Resource Collector | connectionServerHealthResourceCollector | Time taken to complete the collection of Horizon Connection Server Health | ||
| Connection Server Resource Collector | connectionServerResourceCollector | Time taken to complete the collection of Horizon Connection Servers | ||
| Desktop Pool Details Resource Collector | desktopPoolDetailsResourceCollector | Time taken to complete the collection of Horizon Desktop Pool Details | ||
| Desktop Pool List Resource Collector | desktopPoolListResourceCollector | Time taken to complete the collection of Horizon Desktop Pools | ||
| Desktop VM Resource Collector | desktopVMResourceCollector | Time taken to complete the collection of Horizon Desktop VMs | ||
| Horizon Virtual Center Resource Collector | horizonVirtualCenterResourceCollector | Time taken to complete the collection of Horizon Virtual Centres | ||
| Horizon World Resource Collector | horizonWorldResourceCollector | Time taken to complete the collection of Horizon World | ||
| Host or Cluster Resource Collector | hostOrClusterResourceCollector | Time taken to complete the collection of Horizon Hosts or Clusters | ||
| Pod Federation Resource Collector | podFederationResourceCollector | Time taken to complete the collection of Horizon Pod Federations | ||
| Pod Health Resource Collector | podHealthResourceCollector | Time taken to complete the collection of Horizon Pod Health | ||
| Pod Resource Collector | podResourceCollector | Time taken to complete the collection of Horizon Pods | ||
| RDS Farm Details Resource Collector | rDSFarmDetailsResourceCollector | Time taken to complete the collection of Horizon RDS Farm Details | ||
| RDS Farm List Resource Collector | rDSFarmListResourceCollector | Time taken to complete the collection of Horizon RDS Farms | ||
| RDS Host List Resource Collector | rDSHostListResourceCollector | Time taken to complete the collection of Horizon RDS Hosts | ||
| RDS Session Details Resource Collector | rDSSessionDetailsResourceCollector | Time taken to complete the collection of Horizon RDS Session Details | ||
| RDS Session Resource Collector | rDSSessionResourceCollector | Time taken to complete the collection of Horizon RDS Sessions | ||
| Session Resource Collector | sessionResourceCollector | Time taken to complete the collection of Horizon User Sessions | ||
| Site Resource Collector | siteResourceCollector | Time taken to complete the collection of Horizon Sites | ||
| User Details Resource Collector | userDetailsResourceCollector | Time taken to complete the collection of Horizon Users Details | ||
| Users Resource Collector | usersResourceCollector | Time taken to complete the collection of Horizon Users | ||
| vCenter Resources Collector | vcenterResourcesCollector | Time taken to complete the collection of vROPs vCenter Resources | ||
| VDI Session Details Resource Collector | vDISessionDetailsResourceCollector | Time taken to complete the collection of Horizon VDI Session Details | ||
| Virtual Center Health List Resource Collector | virtualCenterHealthListResourceCollector | ime taken to complete the collection of Horizon Virtual Center Health | ||
| Virtual Machine Resource Collector | virtualMachineResourceCollector | Time taken to complete the collection of Horizon Virtual Machine | ||
| Average Application Session Details API Response Time | avgApplicationSessionDetailsApi | Average time taken to receive response from Application Session Details API of Horizon | ||
| Average RDS Session Details API Response Time | avgRDSSessionDetailsApi | Average time taken to receive response from Average RDS Session Details API of Horizon | ||
| Average VDI Session Details API Response Time | avgVDISessionDetailsApi | Average time taken to receive response from Average VDI Session Details API of Horizon | ||
| Application Session Details Resource Collector | applicationSessionDetailsResourceCollector7xx | Time taken to complete the collection of Horizon Application Session Details. | ||
| Application Session Details Resource Collector | applicationSessionDetailsResourceCollector8xx | Average time taken to receive response from Application Session Details API of Horizon | ||
| RDS Session Details Resource Collector | rDSSessionDetailsResourceCollector7xx | Time taken to complete the collection of Horizon RDS Session Details | ||
| RDS Session Details Resource Collector | rDSSessionDetailsResourceCollector8xx | Time taken to complete the collection of Horizon RDS Session Details | ||
| VDI Session Details Resource Collector | vDISessionDetailsResourceCollector8xx | Time taken to complete the collection of Horizon VDI Session Details | ||
| VDI Session Details Resource Collector | vDISessionDetailsResourceCollector7xx | Time taken to complete the collection of Horizon VDI Session Details | ||
| Event Resource Collector | eventResourceCollector | Time taken to complete the collection of Horizon Events | ||
| Event Resource Collector | eventResourceCollector7xx | Time taken to complete the collection of Horizon Events | ||
| Event Resource Collector | eventResourceCollector8xx | Time taken to complete the collection of Horizon Events | ||
| Gateway Resource Collector | gatewayResourceCollector | Time taken to complete the collection of Horizon Gateway Info | ||
| RDS Application Session Vrops Resource Collector | rdsApplicationSessionVropsResourceCollector | Time taken to complete the collection of RDS Application Session resources from VROPS | ||
| RDS Application Session Vrops Stat Collector | rdsApplicationSessionVropsStatCollector | Time taken to complete the collection of RDS Application Session metrics from VROPS | ||
| VDI Application Session Vrops Resource Collector | vdiApplicationSessionVropsResourceCollector | Time taken to complete the collection of VDI Application Session resources from VROPS | ||
| VDI Application Session Vrops Stat Collector | vdiApplicationSessionVropsStatCollector | Time taken to complete the collection of VDI Application Session metrics from VROPS | ||
| Connection Server Vrops Resource Collector | connectionServerVropsResourceCollector | Time taken to complete the collection of Connection Server resources from VROPS | ||
| Connection Server Vrops Stat Collector | connectionServerVropsStatCollector | Time taken to complete the collection of Connection Server metrics from VROPS | ||
| Horizon Adapter Vrops Resource Collector | horizonAdapterVropsResourceCollector | Time taken to complete the collection of Horizon Adapter resources from VROPS | ||
| RDS Farm Vrops Resource Collector | rDSFarmVropsResourceCollector | Time taken to complete the collection of RDS Farm resources from VROPS | ||
| RDS Farm Vrops Stat Collector | rDSFarmVropsStatCollector | Time taken to complete the collection of RDS Farm metrics from VROPS | ||
| RDS Host Vrops Resource Collector | rDSHostVropsResourceCollector | Time taken to complete the collection of RDS Host resources from VROPS | ||
| RDS Host Vrops Stat Collector | rDSHostVropsStatCollector | Time taken to complete the collection of RDS Host metrics from VROPS | ||
| RDS Session Vrops Resource Collector | rDSSessionVropsResourceCollector | Time taken to complete the collection of RDS Session resources from VROPS | ||
| RDS Session Vrops Stat Collector | rDSSessionVropsStatCollector | Time taken to complete the collection of RDS Session metrics from VROPS | ||
| VDI Session Vrops Resource Collector | vDISessionVropsResourceCollector | Time taken to complete the collection of VDI Session resources from VROPS | ||
| VDI Session Vrops Stat Collector | vDISessionVropsStatCollector | Time taken to complete the collection of VDI Session metrics from VROPS | ||
| Virtual Machine Vrops Resource Collector | virtualMachineVropsResourceCollector | Time taken to complete the collection of Virtual Machine resources from VROPS | ||
| Average VDI Application Session Resource Vrops API response time | avgVDIApplicationSessionResourceDtoApi | Average time taken to receive response from VDI Application Session resource API of VROPS | ||
| Average VDI Application Session metrics Vrops API response time | avgVDIApplicationSessionStatApi | Average time taken to receive response from VDI Application Session metrics API of VROPS | ||
| Average RDS Application Session Resource Vrops API response time | avgRDSApplicationSessionResourceDtoApi | Average time taken to receive response from RDS Application Session resource API of VROPS | ||
| Average RDS Application Session metrics Vrops API response time | avgRDSApplicationSessionStatApi | Average time taken to receive response from RDS Application Session metrics API of VROPS | ||
| Average VDI Session Resource Vrops API response time | avgVDISessionResourceDtoApi | Average time taken to receive response from VDI Session resource API of VROPS | ||
| Average VDI Session metrics Vrops API response time | avgVDISessionStatApi | Average time taken to receive response from VDI Session metrics API of VROPS | ||
| Average RDS Session Resource Vrops API response time | avgRDSSessionResourceDtoApi | Average time taken to receive response from RDS Session resource API of VROPS | ||
| Average RDS Session metrics Vrops API response time | avgRDSSessionStatApi | Average time taken to receive response from RDS Session metrics API of VROPS | ||
| Average Virtual machine resource collection Vrops API response time | avgVirtualMachineResourceCollectionPerVCFromVropsApi | Average time taken to receive response from Virtual machine resource API of VROPS | ||
| Average RDS Host Resource Vrops API response time | avgRDSHostResourceDtoApi | Average time taken to receive response from RDS Host Resource API of VROPS | ||
| Average RDS Host metrics Vrops API response time | avgRDSHostStatApi | Average time taken to receive response from RDS Host metrics API of VROPS | ||
| VDI Pool Vrops Resource Collector | vDIPoolVropsResourceCollector | Time taken to complete the collection of VDI Pool resources from VROPS | ||
| VDI Pool Vrops Stat Collector | vDIPoolVropsStatCollector | Time taken to complete the collection of VDI Pool metrics from VROPS | ||
| Event Details Resource Collector | eventDetailsResourceCollector | Time taken to complete the collection of Horizon Event Details | ||
| Event Details Resource Collector | eventDetailsResourceCollector7xx | Time taken to complete the collection of Horizon Event Details | ||
| Event Details Resource Collector | eventDetailsResourceCollector8xx | Time taken to complete the collection of Horizon Event Details | ||
| Average Event Query Response Time | avgEventQueryResponseTime | Average time taken to receive response from Event Query API of Horizon | ||
| UAG Vrops Resource Collector | uAGVropsResourceCollector | Time taken to complete the collection of UAG resources from VROPS | ||
| Performance Collectors/Data Processors/Execution Time | Application Pool Metrics Processor | applicationPoolMetricProcessor | Time taken to complete the processing of Application Pool Metrics | |
| Application Session to Children Relationships Processor | applicationSessionChildrenRelationshipProcessor | Time taken to complete the processing of Application Session to Children Relationships | ||
| Application Session Metrics Processor | applicationSessionMetricProcessor | Time taken to complete the processing of Application Session Metrics | ||
| Application Session to Application Pool Relationships Processor | applicationSessionToApplicationPoolRelationshipProcessor | Time taken to complete the processing of Application Session to Application Pool Relationships | ||
| Connection Server Metrics Processor | connectionServerMetricProcessor | Time taken to complete the processing of Connection Server Metrics | ||
| Connection Server to VM Relationships Processor | connectionServerToVMRelationshipProcessor | Time taken to complete the processing of Connection Servers to VMs Relationships | ||
| Horizon World Relationships Processor | horizonWorldRelationshipProcessor | Time taken to complete the processing of Horizon World Relationships | ||
| Orphaned Entity Metrics Processor | orphanedEntityMetricProcessor | Time taken to complete the processing of Orphaned Entity Metrics | ||
| Pod Federation to Site Relationships Processor | podFederationToSiteRelationshipProcessor | Time taken to complete the processing of Pod Federation to Site Relationships Processor | ||
| Pod Metrics Processor | podMetricProcessor | Time taken to complete the processing of Pod Metrics | ||
| Pod to Connection Server Relationships Processor | podToConnectionServerRelationshipProcessor | Time taken to complete the processing of Pod to Connection Server Relationships | ||
| Pod to vCenter Adapter Instance Relationships Processor | podToVCenterAdapterInstanceRelationshipProcessor | Time taken to complete the processing of Pod to vCenter Adapter Instance Relationships | ||
| RDS Desktop Pool Metrics Processor | rDSDesktopPoolMetricProcessor | Time taken to complete the processing of RDS Desktop Pool Metrics | ||
| RDS Farm Metrics Processor | rDSFarmMetricProcessor | Time taken to complete the processing of RDS Farm Metrics | ||
| RDS Farm Relationships Processor | rDSFarmRelationshipProcessor | Time taken to complete the processing of RDS Farm Relationships | ||
| RDS Farm to Application Pool Relationships Processor | rDSFarmToApplicationPoolRelationshipProcessor | Time taken to complete the processing of RDS Farm to Application Pool Relationships | ||
| RDS Farm to Pool Relationships Processor | rDSFarmToPoolRelationshipProcessor | Time taken to complete the processing of RDS Farm to Pool Relationships | ||
| RDS Farm to Server Relationships Processor | rDSFarmToServerRelationshipProcessor | Time taken to complete the processing of RDS Farm to Server Relationships | ||
| RDS Host Metrics Processor | rDSHostMetricProcessor | Time taken to complete the processing of RDS Host Metrics | ||
| RDS Host to VM Relationships Processor | rDSHostToVMRelationshipProcessor | Time taken to complete the processing of RDS Host to VM Relationships | ||
| RDS Pool to RDS Session Relationships Processor | rDSPoolToRDSSessionRelationshipProcessor | Time taken to complete the processing of RDS Pool to RDS Session Relationships | ||
| RDS Session Metrics Processor | rDSSessionMetricProcessor | Time taken to complete the processing of RDS Session Metrics | ||
| RDS Session to RDS Host Relationships Processor | rdsSessionToRdsHostRelationshipProcessor | Time taken to complete the processing of RDS Session to RDS Host Relationships | ||
| Site to Pod Relationships Processor | siteToPodRelationshipProcessor | Time taken to complete the processing of Site to Pod Relationships | ||
| User Metrics Processor | userMetricProcessor | Time taken to complete the processing of User Metrics | ||
| User to Application Session Relationships Processor | userToApplicationSessionRelationshipProcessor | Time taken to complete the processing of User to Application Session Relationships | ||
| User to Entitlement Relationships Processor | userToEntitlementRelationshipProcessor | Time taken to complete the processing of User to Entitlement Relationships | ||
| User to RDS Session Relationships Processor | userToRDSSessionRelationshipProcessor | Time taken to complete the processing of User to RDS Session Relationships | ||
| VDI Pool Metrics Processor | vDIPoolMetricProcessor | Time taken to complete the processing of VDI Pool Metrics | ||
| VDI Pool Relationships Processor | vDIPoolRelationshipProcessor | Time taken to complete the processing of VDI Pool Relationships | ||
| VDI Pool to Application Pool Relationships Processor | vDIPoolToApplicationPoolRelationshipProcessor | Time taken to complete the processing of VDI Pool to Application Pool Relationships | ||
| VDI Pool to Entitlement Relationships Processor | vDIPoolToEntitlementProcessor | Time taken to complete the processing of VDI Pool to Entitlement Relationships | ||
| VDI Pool to VM Relationships Processor | vDIPoolToVMRelationshipProcessor | Time taken to complete the processing of VDI Pool To VM Relationships | ||
| VDI Session Metrics Processor | vDISessionMetricProcessor | Time taken to complete the processing of VDI Session Metrics | ||
| VDI Session to VM Relationships Processor | vDISessionToVMRelationshipProcessor | Time taken to complete the processing of VDI Session to VM Relationships | ||
| Application Session Metrics Processor | applicationSessionMetricProcessor7xx | Time taken to complete the processing of Application Session Metrics | ||
| Application Session Metrics Processor | applicationSessionMetricProcessor8xx | Time taken to complete the processing of Application Session Metrics | ||
| RDS Session Metrics Processor | rDSSessionMetricProcessor7xx | Time taken to complete the processing of RDS Session Metrics | ||
| RDS Session Metrics Processor | rDSSessionMetricProcessor8xx | Time taken to complete the processing of RDS Session Metrics | ||
| VDI Session Metrics Processor | vDISessionMetricProcessor7xx | Time taken to complete the processing of VDI Session Metrics | ||
| VDI Session Metrics Processor | vDISessionMetricProcessor8xx | Time taken to complete the processing of VDI Session Metrics | ||
| UAG to Application Session Relationship Processor | uAGToApplicationSessionRelationshipProcessor | Time taken to complete the processing of UAG to Application Session Relationships | ||
| UAG to Connection Server Relationship Processor | uAGToConnectionServerRelationshipProcessor | Time taken to complete the processing of UAG to Connection Server Relationships | ||
| UAG to RDS Session Relationship Processor | uAGToRDSSessionRelationshipProcessor | Time taken to complete the processing of UAG to RDS Session Relationships | ||
| UAG to VDI Session Relationship Processor | uAGToVDISessionRelationshipProcessor | Time taken to complete the processing of UAG to VDI Session Relationships | ||
| UAG to VM Relationship Processor | uAGToVMRelationshipProcessor | Time taken to complete the processing of UAG to VM Relationships | ||
| UAG Metric Processor | uAGMetricProcessor | Time taken to complete the processing of UAG Metrics |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| VDI Pool | Capacity | Number of Available Desktops | available | The number of available desktops in a ready to connect to state. If this number reaches 0, additional desktops will need to be created to support more user sessions |
| Number of Connected Desktops | connected | Number of Connected Desktop VMs | ||
| Number of Disconnected Desktops | disconnected | Number of DisConnected Desktop VMs | ||
| Number of VDI Sessions | vdiSessions | Number of active VDI Sessions in the Pool | ||
| Application Session Count | appSessions | Number of active Application Sessions in the Pool. | ||
| Number of Spare Desktops | spareDesktops | Number of Spare Desktops | ||
| Usable Capacity | usableCapacity | Capacity Available in Percentage | ||
| Used Capacity | usedCapacity | The percentage of desktops in a VDI Pool that are considered in use out of the total number of desktops. If Capacity Used = 100%, no additional desktops are available | ||
| Lowest Available Desktops | minAvailableDesktops | Lowest number of available desktops in a 30 day period. Can be used to determine if the configured pool desktop capacity is sufficient | ||
| Maximum Desktops Used | maxConnectedDesktops | Maximum number of concurrent desktops in use in a 30 day period. Can be compared against the Lowest Available Desktops count to determine if the configured pool desktop capacity is sufficient | ||
| Datastore | Available Capacity (GB) | freeMemory | Free Capacity of the Datastore in GB | |
| Availability | Desktops in Bad State | noOfBadStateDesktops | Sum of all desktops in the pool that is not in ready or used state. Desktops that have problem should be 0 most of the time | |
| Summary | Number Of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources | |
| Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state | ||
| Number of Disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to | ||
| Performance | 95th Percentile Datacenter | dataCenter95thPercentile | 95th Percentile Datacenter KPI from all the VDIPool based Sessions | |
| 95th Percentile Network | network95thPercentile | 95th Percentile Network KPI from all the VDIPool based sessions | ||
| Session | Worst time taken to load profile | worstLoadProfileTime | The longest time taken to load user profile among all the user sessions | |
| Worst time taken to Logon | worstLogOnTime | The longest time taken to login to the desktop among all users in the VDI Pool | ||
| CPU | Sessions with CPU Queue Length | sessionCpuQueueLength | The percentage of sessions with CPU Queue Length above 2 per vCPU | |
| Sessions with CPU Ready | sessionCpuReady | The percentage of sessions with CPU Ready above 2.5% | ||
| Worst CPU Co-stop | highestCpuCostop | Highest CPU Co-stop from all sessions. Keep this number below 2.5% | ||
| Worst CPU Queue Length | highestCpuQueueLength | Highest CPU Queue from all active sessions. Keep this number below 3 queue per vCPU | ||
| Worst CPU Ready | highestCpuReady | Highest CPU Ready from VDI Sessions. Keep this number below 2.5% | ||
| Number of Sessions with High CPU Utilization | sessionsCpuUtilizationAboveThreshold | Count of Sessions where CPU Usage > 95%. Expect this number to be minimal. Note that high performance typically requires high utilization | ||
| Utilization | utilization | The total CPU Usage in Ghz of all the VMs in a VDI Pool | ||
| Total Capacity | totalCapacity | The size of the VDI Pool capacity in terms of CPU resources | ||
| Worst Context Switch | worstContextSwitch | The worst CPU Context Switch among all the RDS Hosts in the farm. Ensure this is within the threshold you see during baselining exercise | ||
| Active Session CPU Utilization | activeSessionutilization | CPU Utilization calculated from the underlying active Sessions | ||
| Memory | Sessions with Memory Contention | sessionMemoryContention | The percentage of sessions with Memory Contention above 1% | |
| Worst Memory Contention | highestMemoryContention | Highest Memory Contention from all Sessions. Keep this number below 1% | ||
| Lowest free Memory among Sessions | lowestFreeMemory | Lowest free Memory among Sessions | ||
| Worst Memory Page In Rate | highestMemoryPageInRate | Highest Memory page-in rate from all active sessions | ||
| Number of Sessions with Low Available Memory | sessionFreeMemorybelowThreshold | Count of Sessions with Available Memory < 500 MB. Expect this number to be minimal for ideal performance | ||
| Utilization | utilization | The total memory used in GB across all of the VMs in the VDI Pool | ||
| Total Capacity | totalCapacity | The size of the VDI Pool capacity in terms of memory resources | ||
| Protocol | Sessions with Protocol Latency | sessionProtocolLatency | The percentage of sessions with Protocol Latency above 50 ms | |
| Sessions with Protocol transmit dropped packet | sessionPacketLossTransmit | The percentage of sessions with Protocol transmit dropped packet above 0.5% | ||
| Sessions with Protocol receive dropped packet | sessionPacketLossReceive | The percentage of sessions with Protocol receive dropped packet above 0.5% | ||
| Lowest Frame rate | lowestFrameRate | The lowest number among all the active sessions frame rate counter. A low frame rate result in inferior user experience. Used this to track if any of the session is running on low frame rate. While occasional low is fine, a prolonged low could lead to degraded user experience | ||
| Average Frame Rate | averageFrameRate | The average of frame rate of all active sessions. A low frame rate result in inferior user experience | ||
| Worst Protocol Latency | highestLatency | The percentage of sessions with Protocol Latency above 50 ms | ||
| Worst Protocol Transmit Packet Loss | highestPacketLossTransmit | Highest Horizon Protocol Transmitted Packet Loss from all active sessions | ||
| Worst Protocol Receive Packet Loss | highestPacketLossReceive | Highest Horizon Protocol Received Packet Loss from all active sessions | ||
| Number of Sessions with High Protocol Latency | sessionProtocolLatencyAboveThreshold | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) latency > 180 ms. Expect this number to be low. In ideal state it is 0 | ||
| Number of Sessions with High Protocol Packet Receive Loss | sessionPacketLossReceiveAboveThreshold | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) packet transmit loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Number of Sessions with High Protocol Packet Transmit Loss. | sessionPacketLossTransmitAboveThreshold | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) packet receive loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Average Bandwidth Transmit Utilization | avgTransmittedBandwidth | The average of bandwidth transmit utilization of all active sessions in the farm. Sessions include both Desktop and App Sessions. Use this number as input for bandwidth capacity planning | ||
| Total Bandwidth Transmit Utilization | totalTransmittedBandwidth | he total of bandwidth transmit utilization of all active sessions in the farm. Sessions include both Desktop and App Sessions. Use this number as input for bandwidth capacity planning | ||
| Disk | Sessions with Disk Latency | sessionDiskLatency | The percentage of sessions with Disk Latency above 20 ms% | |
| Sessions with Disk Queue Length | sessionDiskQueueLength | The percentage of sessions with Disk Queue Length above 10 | ||
| Worst vDisk Latency | highestDiskLatency | Highest VM virtual disk latency from all active Session. Aim for this number to be below 10 ms. | ||
| Worst Disk Queue | highestDiskQueueLength | Highest Microsoft Windows Disk Queue from all active Sessions. The queue length indicates the number of IO that are not yet processed | ||
| Number of Sessions with High Disk Latency | sessionDiskLatencyAboveThreshold | Count of Sessions with Disk Latency > 15 ms. Expect this number to be minimal, especially on SSD-backed storage | ||
| Disk IOPS | dsikIops | Sum of all VDI Pools I/O operations per second. It includes both Reads and Writes | ||
| Read Disk IOPS | diskReadIops | Sum of all VDI Pools I/O operations per second. It includes only Read operations | ||
| Number of Write Disk IOPS | diskWriteIops | Sum of all VDI Pools I/O operations per second. It includes only Write operations | ||
| Disk Throughput | diskThroughput | Sum of all VDI Pools disk throughput. It includes both Reads and Writes | ||
| Read Disk Throughput | diskReadThroughput | Sum of all VDI Pools disk throughput. It includes only Reads | ||
| Write Disk Throughput | diskWriteThroughput | Sum of all VDI Pools disk throughput. It includes only Writes | ||
| Worst Outstanding IO | worstOutstandingIO | The highest disk outstanding IO among all the sessions in the pool. Keep this number below best practice | ||
| Worst vDisk Read Latency | peakvDiskReadLatency | Highest VM virtual disk read latency from all active sessions | ||
| Worst vDisk Write Latency | peakvDiskWriteLatency | Highest VM virtual disk write latency from all active sessions | ||
| Performance | Datacenter KPI | datacenter | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the datacenter segment, as opposed to the network. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | |
| Network KPI | network | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the network segment of session, as opposed to the datacenter. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | ||
| Worst datacenter among Sessions | worstDatacenter | Lowest Datacenter KPI from all sessions. Aim for this number to be in the green range (75% - 100%) | ||
| Number of Sessions with Red DatacenterKPI | redDatacenterCount | he number of Sessions with Datacenter KPI value falling within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of Sessions with Red Network KPI | redNetworkCount | The number of Sessions with Network KPI value falling within the red range (0% - 25%). Expect this number to be 0 | ||
| Lowest Network KPI among Sessions | sessionLowestNetwork | Lowest Network KPI among Sessions | ||
| Summary/Oversized | Memory | memory | Values that represent how much is the resources like cpu, memory etc oversized | |
| Virtual CPUs | vCpus | Number of additional vCPUs used than Allocated | ||
| Over Sized Users | users | Number of users having over-sized Memory or vCPUs | ||
| Summary/Undersized | Memory | memory | Values that represent how much is the resources like cpu, memory etc undersized | |
| Virtual CPUs | vCpus | Number of additional vCPUs required on this pool for this User | ||
| Under Sized Users | users | Number of users having under-sized Memory or vCPUs | ||
| Configuration | Desktop Memory Size | memoryDesktopSize | Average Memory size of a Desktop in the VDI Pool | |
| Desktop CPU Size | cpuDesktopSize | Average CPU size of a Desktop in the VDI Pool |
| Resrouce name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| RDS Desktop Tool | Summary | Number of Sessions | totalSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources |
| Resrouce name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| RDS FarmRDS Farm | Summary | Sessions per RDS Host | noOfSessionsPerRDSHost | The average number of sessions per host in the farm. Compare this with the plan. If the reality differs with the plan, adjust accordingly |
| Datastore (Only for Automated Farm) | Available Capacity (GB) | freeMemory | Free Capacity of the Datastore in GB | |
| Capacity | Number of Sessions Remaining | sessionRemaining | Number of additional sessions that can be created on the Farm | |
| 99th Percentile of Number of Sessions Remaining | sessionRemaining99thPercentile | 99th Percentile of number of sessions remaining considering sample data for a month | ||
| Average Session CPU Usage | sessionOverCommittedCpu | The average among all the RDS Session's CPU utilization. Used in capacity management | ||
| Average Session Memory Usage | sessionOverCommittedMem | The average among all the RDS Session's Memory utilization. Used in capacity management | ||
| Number of Sessions Over Committed on CPU | sessionSizeCpu | Number of Sessions Over Committed on CPU | ||
| Number of Sessions Over Committed on Memory | sessionSizeMemory | Number of Sessions Over Committed on Memory | ||
| RDS Host Remaining | rdshostRemaining | RDS Host Remaining | ||
| Total CPU Usage | farmUsageCpuMHz | Total CPU Usage | ||
| Total Memory Usage | farmUsageMemoryMB | Total Memory Usage | ||
| Performance | 95th Percentile Network KPI | network95thPercentile | 95th Percentile Network KPI from all the VDIPool and Farms | |
| 95th Percentile Datacenter KPI | dataCenter95thPercentile | 95th Percentile Datacenter KPI from all the VDIPool and Farms | ||
| CPU Usage per RDS Host | avgCPUUsagePerRdsHost | The average Usage of CPU per host in the farm | ||
| Memory Usage per RDS Host | avgMemoryUsagePerRdsHost | he average Usage of Memory per host in the farm | ||
| Availability | Number of RDS Hosts in Bad State | noOfBadStateRDSHosts | Sum of all RDS Hosts in the farm that is not in ready or used state. RDS Hosts that have problem should be 0 most of the time | |
| Session | Worst time taken to load profile | worstLoadProfileTime | The longest time taken to load user profile among all the user sessions in the Farm | |
| Worst time taken to Logon | worstLogOnTime | The longest time taken to login among all the user sessions in the Farm | ||
| CPU | Peak CPU Utilization among RDS Hosts | peakUtilization | Peak CPU Utilization among RDS Hosts | |
| Usage disparity among RDS Hosts | usageDisparity | CPU Usage Disparity among RDS Hosts | ||
| Worst CPU Co-stop among RDS Hosts | highestCpuCostop | Highest CPU Co-stop from all RDS Hosts. Keep this number below 2.5% | ||
| Worst CPU Queue Length among RDS Hosts | highestCpuQueueLength | Highest CPU Queue from all RDS Hosts. Keep this number below 3 queue per vCPU | ||
| Worst CPU Ready among RDS Hosts | highestCpuReady | Highest CPU Ready from all RDS Hosts. Keep this number below 2.5% | ||
| Number of RDS Hosts with high CPU utilization | noOfRdsHostWithHighCPUUsage | Count of RDS Hosts with CPU Usage > 95%. Expect this number to be minimal. Note that high performance typically requires high utilization | ||
| Total Utilization | utilization | The total CPU Usage in Ghz of all the VMs in the RDS Farm | ||
| Worst Context Switch | worstContextSwitch | he worst CPU Context Switch among all the RDS Hosts in the farm. Ensure this is within the threshold you see during baselining exercise | ||
| Memory | Lowest free Memory among RDS Hosts | lowestFreeMemory | Lowest free Memory among RDS Hosts | |
| Usage disparity among RDS Hosts | usageDisparity | Usage disparity among RDS Hosts | ||
| Worst Memory Contention | highestMemoryContention | Highest Memory Contention from all RDS Hosts. Keep this number below 1% | ||
| Worst Page-in rate | highestMemoryPageInRate | Highest Memory page-in rate from all active RDS Hosts | ||
| Number of RDS Hosts with low available Memory | noOfRdsHostWithLowestFreeMemory | Count of RDS Hosts with Available Memory < 500 MB. Expect this number to be minimal for ideal performance | ||
| Total Utilization | utilization | The total memory used in GB across all of the VMs in the RDS Farm | ||
| Disk | Worst vDisk Latency | highestDiskLatency | Highest VM virtual disk latency from all active RDS Hosts. Aim for this number to be below 10 ms. | |
| Read Disk IOPS | dsikReadIops | Sum of all RDS Farms I/O operations per second. It includes only Read operations | ||
| Write Disk IOPS | diskWriteIop | Sum of all RDS Farms I/O operations per second. It includes only Write operations | ||
| Read Disk Throughput | diskReadThroughput | Sum of all RDS Farms disk throughput. It includes only Reads | ||
| Write Disk Throughput | diskWriteThroughput | Sum of all RDS Farms disk throughput. It includes only Writes | ||
| Worst Disk Queue | highestDiskQueueLength | Highest Microsoft Windows Disk Queue from all active RDS Hosts. The queue length indicates the number of IO that are not yet processed | ||
| Disk IOPS | diskIOPS | Sum of all RDS I/O operations per second. It includes both Reads and Writes. Investigate if this becomes excessive. | ||
| Disk Throughput | diskThroughPut | Sum of Disk Throughput of all RDS Hosts in the farm. | ||
| Number of RDS Hosts with high Disk latency | noOfRdsHostWithHighDiskLatency | Count of RDS Hosts with Disk Latency > 15 ms. Expect this number to be minimal, especially on SSD-backed storage | ||
| Worst Outstanding IO | worstOutstandingIO | The highest disk outstanding IO among all the RDS Hosts. Keep this number below best practice | ||
| Worst vDisk Read Latency | peakvDiskReadLatency | Highest read latency among any of the VM virtual disks. Keep this number below 10 ms | ||
| Worst vDisk Write Latency | peakvDiskWriteLatency | Highest write latency among any of the VM virtual disks. Keep this number below 10 ms | ||
| Protocol | Lowest Frame rate | lowestFrameRate | The lowest number among all the active sessions frame rate counter. A low frame rate result in inferior user experience. Used this to track if any of the session is running on low frame rate. While occasional low is fine, a prolonged low could lead to degraded user experience | |
| Average Frame rate | averageFrameRate | The average of frame rate of all active sessions. A low frame rate result in inferior user experience | ||
| Worst Protocol Latency | highestLatency | Highest Horizon Protocol latency from all active sessions. A user can have >1 sessions concurrently. Keep this number below 180 ms | ||
| Worst Protocol Transmit Packet Loss | highestPacketLossTransmit | Highest Horizon Protocol Transmitted Packet Loss from all active sessions | ||
| Worst Protocol Receive Packet Loss | highestPacketLossReceive | Highest Horizon Protocol Received Packet Loss from all active sessions | ||
| Number of RDS Sessions with High Protocol Latency | sessionLatencyAboveThreshold | Count of RDS Sessions with Horizon protocol (e.g. Blast, PCoIP) latency > 180 ms. Expect this number to be low. In ideal state it is 0. | ||
| Number of RDS Sessions with High Protocol Packet Transmit Loss | sessionPacketLossTransmitAboveThreshold | Count of RDS Sessions with Horizon protocol (e.g. Blast, PCoIP) packet transmit loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Number of RDS Sessions with High Protocol Packet Receive Loss | sessionPacketLossReceiveAboveThreshold | Count of RDS Sessions with Horizon protocol (e.g. Blast, PCoIP) packet receive loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Average Bandwidth Transmit Utilization | avgTransmittedBandwidth | The average of bandwidth transmit utilization of all active sessions in the farm. Sessions include both Desktop and App Sessions. Use this number as input for bandwidth capacity planning | ||
| Total Bandwidth Transmit Utilization | totalTransmittedBandwidth | The total of bandwidth transmit utilization of all active sessions in the farm. Sessions include both Desktop and App Sessions. Use this number as input for bandwidth capacity planning | ||
| Performance | Datacenter KPI | datacenter | Average Datacenter KPI from all the RDS Hosts in the farm. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | |
| Network KPI | network | Average Network KPI from all the RDS and Application Sessions from Farm. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | ||
| Worst Datacenter KPI among Sessions | worstDatacenter | Lowest Datacenter KPI from all active RDS Hosts. Aim for this number to be in the green range (75% - 100%) | ||
| Number of Sessions with Red Datacenter KPI | redDatacenterCount | The number of RDS Host with Datacenter KPI value falling within the red range (0% - 25%). Expect this number to be 0 | ||
| Lowest Network KPI among Sessions | sessionLowestNetwork | Lowest Network KPI among Sessions | ||
| Number of RDS Hosts with Red Datacenter KPI | redNetworkCount | The number of RDS Host with Datacenter KPI value falling within the red range (0% - 25%). Expect this number to be 0 | ||
| Summary | Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state on RDS Host | |
| umber of Disconnected Sessions | disconnectedSessions | Number of sessions at present that are in disconnected state on RDS Host | ||
| Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources | ||
| RDS Hosts | Agent Unreachable | agentUnreachable | Number of Desktops failed with Agent Unreachable Message | |
| Provisioning Error | provisioningError | An error occurred during provisioning | ||
| RDSH hosts in error states | noOfBadStateRDSHosts | RDSH hosts in error states |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Application Pool | Summary | Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources. |
| CPU | Average Utilization | avgUtilization | The average number of all the session's CPU usage. |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| RDS HostRDSHost | Utilization | Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources. |
| Summary | Number of Connected Sessions | noOfConnectedSessions | Number of sessions at present that are in connected state on RDS Host. | |
| Number of Disconnected Sessions | noOfDisconnectedSessions | Number of sessions at present that are in disconnected state on RDS Host. | ||
| RDS Host Session Disparity | sessionDisparity | The % difference of RDS Host session count against the RDS Farms "RDS Sessions per Host" average. Disparity > 30% indicates the RDS host has an abnormal workload and requires further analysis as to the cause. | ||
| Performance | CPU Usage disparity among RDS Hosts | cpuUsageDisparity | The % difference between the Farm average RDS Host CPU % workload and the individual RDS Host CPU % workload. Disparity > 30% indicates the RDS host has an abnormal workload and requires further analysis as to the cause. | |
| Memory Usage disparity among RDS Hosts | memoryUsageDisparity | The % difference between the Farm average RDS Host Memory workload and the individual RDS Host Memory workload. Disparity > 30 % indicates the RDS Host has an abnormal workload and requires further analysis as to the cause. | ||
| CPU | Co-stop | costopAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Co-Stop among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time the VM is ready to run, but is unable due to co-scheduling constraints. VM with less vCPU have lower co-stop value. | |
| Ready | readyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Ready among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of CPU the VM is ready to run, but unable due to ESXi has no ready physical core to run it. High Ready value impacts VM performance. | ||
| IO Wait | iowaitAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU IO Wait among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time VM CPU is waiting for IO. Formula is Wait - Idle - Swap Wait. High value indicates slow storage subsystem. | ||
| Usage | usageAvg | CPU Usage divided by VM CPU Configuration in MHz. | ||
| Run Queue | cpuqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Run Queue among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of ready threads queuing in the CPU. A number greater than 2 for prolong period indicates CPU core bottleneck | ||
| Utilization | utilization | Amount of actively used virtual CPU. This is the host's view of the CPU usage not the guest operating system view. | ||
| Configured Capacity | totalCapacity | Configured Capacity in GHz, based on nominal (static) frequency of the CPU | ||
| Context Switch | contextSwapRateLatest | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Context Switch Rate, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the rate at which Operating System processes running in a CPU are loaded and unloaded. As context switch is an overhead cost, a high number will impact the application performance. The number varies per application so profile your environment to baseline the expected range. | ||
| Overlap | overlapSummation | Percentage of time where the VM was interrupted by hypervisor as it needs to perform system services on behalf of that VM or other VMs. | ||
| Memory | Available Memory | freeMemoryAvg | Available memory on the guest OS computed as a sum of guest memory standby core, guest memory standby normal, guest memory standby reserve and guest memory free counters from the virtual machine. | |
| Page In Rate | pageInRatePerSecondAvg | The rate at which memory pages are paged in. An unexpected high value could indicate Guest OS does not have enough cache. This can cause performance problem for memory intensive application. | ||
| Contention | contentionAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Memory Contention, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the amount of time, in percentage, the VM CPU is waiting for memory to be brought in. Keep this number below 1%. | ||
| Utilization | utilization | Amount of memory utilized by the Virtual Machine. Reflects the guest OS memory required (for certain vSphere and VMTools versions) or Virtual Machine consumption. | ||
| Configured Capacity | totalCapacity | Memory resources allocated to the Virtual Machine. | ||
| Page Out Rate per second | pageOutRatePerSecondAvg | The rate at which memory pages are paged out. An unexpected high value could indicate Guest OS is under memory pressure. It's a capacity, not performance metric. | ||
| Disk | Disk Queue | diskqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Disk Queue Length, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of outstanding requests + IO currently in progress. | |
| Latency | totalLatencyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest latency among any of the virtual disks, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the weighted average of read latency and write latency. | ||
| Disk IOPS | diskIops | Number of read/write operations per second. This is averaged over the reporting period. | ||
| Read Disk IOPS | diskReadIops | Number of read operations per second. This is averaged over the reporting period. | ||
| Write Disk IOPS | diskWriteIops | Number of write operations per second. This is averaged over the reporting period. | ||
| Disk Throughput | diskThroughput | Amount of data read from/written to storage in a second. This is averaged over the reporting period. | ||
| Read Disk Throughput | diskReadThroughput | Amount of data read from storage in a second. This is averaged over the reporting period. | ||
| Write Disk Throughput | diskWriteThrouhgput | Amount of data written to storage in a second. This is averaged over the reporting period. | ||
| Worst vDisk Read Latency | peakvDiskReadLatency | Highest read latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Worst vDisk Write Latency | peakvDiskWriteLatency | Highest write latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Outstanding IO | vDiskOIO | Amount of disk Input or Output commands waiting in the queue to be executed. High IO, coupled with high latency, impacts performance. This number should be below the queue depth of the OS. | ||
| Performance | Datacenter KPI | datacenter | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the datacenter segment, as opposed to the network. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Horizon Pod | Utilization | Number of Connected Users | connectedUsers | Number of users who have atleast one connected session at present |
| Number of Users | totalUsers | The number of users who have at least one connected or disconnected session at present | ||
| Number of DisConnected Users | disconnectedUsers | Number of users who have only disconnected sessions | ||
| Performance | 95th Percentile Datacenter KPI | dataCenter95thPercentile | 95th Percentile Datacenter KPI from all VDI Based Pools | |
| 95th Percentile UAG Servers Datacenter KPI | network95thPercentile | 95th Percentile UAG Servers Datacenter KPI | ||
| 95th Percentile Connection Servers Datacenter KPI | csDataCenter95thPercentile | 95th Percentile Connection Servers Datacenter KPI | ||
| 95th Percentile Farms Datacenter KPI | farmDataCenter95thPercentile | 95th Percentile Farms Datacenter KPI | ||
| CPU Usage per Connection Server | avgCPUUsagePerCS | The average Usage of CPU per Connection server in the Horizon Pod | ||
| Memory Usage per Connection Server | avgMemoryUsagePerCS | The average Usage of Memory per Connection server in the Horizon Pod | ||
| 95th Percentile Farms Network KPI | farmNetwork95thPercentile | 95th Percentile Farms Network KPI | ||
| 95th Percentile UAG Servers Datacenter KPI | uagDataCenter95thPercentile | 95th Percentile UAG Servers Datacenter KPI | ||
| Available Capacity | freeMemory | Free Capacity of the Datastore in GB | ||
| Number of Unhealthy Connection Servers | unHealthyConnectionServers | Number of Unhealthy Connection Servers under the Pod | ||
| Connections per Connection Server | noOfConnectionsPerCS | Average Connections per Connection Server used across the Horizon Pod. This should be compared against the Connection Disparity metric of Connection Servers. | ||
| Number of VDI Pools | vdiPoolCount | The number of VDI Pools in the inventory at present | ||
| Number of RDS Desktop Pools | rdsDesktopPoolCount | The number of RDS Desktop Pools in the inventory at present | ||
| Number of Application Pools | applicationPoolCount | The number of Application Pools in the inventory at present | ||
| Session | Worst time taken to load profile | worstLoadProfileTime | The longest time taken to load user profile among all the user sessions. | |
| Worst time taken to Logon | worstLogOnTime | The longest time taken to login to the desktop among all users in the VDI Pool | ||
| Performance | Datacenter KPI | datacenter | Average Datacenter KPI from all the RDS Hosts,VDI Sessions and Application Sessions from VDI Pools in the Pod. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | |
| Network KPI | network | Average Network KPI from all the VDI, RDS and Application Sessions in the Pod. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). | ||
| Number of VDI Sessions with Red Network KPI | vdiSessionsWithRedNetworkKPI | The number of VDI Sessions with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of VDI Sessions with Red Datacenter KPI | vdiSessionsWithRedDatacenterKPI | The number of VDI Sessions with Datacenter KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of VDI Pools with Red Network KPI | vdiPoolsWithRedNetworkKPI | The number of VDI Pools with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of VDI Pools with Red Datacenter KPI | vdiPoolsWithRedDatacenterKPI | The number of VDI Pools with Datacenter KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of Application Sessions with Red Network KPI | appSessionsWithRedNetworkKPI | The number of Application Sessions with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of RDS Sessions with Red Network KPI | rdsSessionsWithRedNetworkKPI | The number of RDS Sessions (App or Desktop) with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Configuration | Number of Desktops in bad state | noOfBadStateDesktops | Sum of all desktops in the pool that is not in ready or used state. Desktops that have problem should be 0 most of the time. | |
| Number of RDS Hosts in Bad State | noOfBadStateRDSHosts | Sum of all RDS Hosts in the farm that is not in ready or used state. RDS Hosts that have problem should be 0 most of the time | ||
| Disk | Number of VDI Sessions with High Disk Latency | vdiSessionsWithHighCPUUtilization | Count of VDI Sessions with Disk Latency > 15 ms. Expect this number to be minimal, especially on SSD-backed storage. | |
| Disk IOPS | diskIops | Sum of all VDI VMs and RDS Hosts I/O operations per second. It includes both Reads and Writes. Investigate if this becomes excessive. | ||
| Read Disk IOPS | diskReadIops | Sum of all VDI VMs and RDS Hosts I/O operations per second. It includes only Read operations. Investigate if this becomes excessive | ||
| Write Disk IOPS | diskWriteIops | Sum of all VDI VMs and RDS Hosts I/O operations per second. It includes only Write operations. Investigate if this becomes excessive | ||
| Disk Throughput | diskThroughput | Sum of all VDI VMs and RDS Hosts disk throughput. It includes both Reads and Writes. Investigate if this becomes excessive | ||
| Read Disk Throughput | diskReadThroughput | Sum of all VDI VMs and RDS Hosts disk throughput. It includes only Reads. Investigate if this becomes excessive | ||
| Write Disk Throughput | diskWriteThroughput | Sum of all VDI VMs and RDS Hosts disk throughput. It includes only Writes. Investigate if this becomes excessive | ||
| Memory | Number of VDI Sessions with Low Available Memory | vdiSessionsWithLowAvailableMemory | Count of VDI Sessions with Available Memory < 500 MB. Expect this number to be minimal for ideal performance. | |
| Utilization | utilization | Sum of all VDI VMs and RDS Hosts Memory Utilization | ||
| Total Capacity | totalCapacity | Sum of all VDI VMs and RDS Hosts capacity that are part of Horizon | ||
| CPU | Number of VDI Sessions with High CPU Utilization | vdiSessionsWithHighCPUUtilization | Count of VDI Sessions where CPU Usage > 95%. Expect this number to be minimal. Note that high performance typically requires high utilization | |
| Utilization | utilization | Sum of all VDI VMs and RDS Hosts CPU Utilization | ||
| Total Capacity | totalCapacity | Sum of all VDI VMs and RDS Hosts capacity that are part of Horizon | ||
| Protocol | Number of Sessions with High Protocol Latency | sessionsWithHighProtocolLatency | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) latency > 180 ms. Expect this number to be low. In ideal state it is 0 | |
| Number of Sessions with High Protocol Packet Loss Transmit | sessionsWithHighProtocolPacketLossTransmit | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) packet transmit loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Number of Sessions with High Protocol Packet Loss Receive | sessionsWithHighProtocolPacketLossReceive | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) packet receive loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Summary | Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources | |
| Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state | ||
| Number of disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to. | ||
| Number of Connection Servers Enabled | noOfConnectionServersEnabled | Number of Connection Servers Enabled | ||
| Number of Connection Servers | connectionServerCount | The number of Horizon Connection Servers in the Pod. It counts any server regardless of operational status as long as it is registered with the Pod | ||
| Number of Managed vCenter Servers | managedVcServersCount | Number of vCenter Servers hosting the Horizon Pod |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Cloud Pod Federation | Summary | Number of VDI Pools | vdiPoolCount | The number of VDI Pools in the Cloud Pod Federation at present |
| Number of RDS Desktop Pools | rdsDesktopPoolCount | The number of RDS Desktop Pools in the Cloud Pod Federation at present | ||
| Number of Application Pools | applicationPoolCount | The number of Application Pools in the Cloud Pod Federation at present | ||
| Number of Pods | podCount | The present number of pods in the Cloud Pod | ||
| Number of Sites | siteCount | Number of Sites | ||
| Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources | ||
| Number of Disconnected Sessions | disConnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to | ||
| Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state | ||
| Number of VDI Machines | vdiVirtualMachineCount | The number of VDI Virtual Machines in the Pod at present | ||
| Performance | Datacenter KPI | datacenter | Average Datacenter KPI from all the RDS Hosts,VDI Sessions and Application Sessions from VDI Pools in the Pod. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | |
| Network KPI | network | Average Network KPI from all the VDI, RDS and Application Sessions in the Pod. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | ||
| Utilization | Number of Connected Users | connectedUsers | Number of users who have atleast one connected session at present | |
| Number of Disonnected Users | disconnectedUsers | Number of users who have only disconnected sessions | ||
| Number of Users | totalUsers | The number of users who have at least one connected or disconnected session at present | ||
| Configuration | Number of Desktops in Bad State | noOfBadStateDesktops | Sum of all desktops in the pool that is not in ready or used state. Desktops that have problem should be 0 most of the time | |
| Number of RDS Hosts in Bad State | noOfBadStateRDSHosts | Sum of all RDS Hosts in the farm that is not in ready or used state. RDS Hosts that have problem should be 0 most of the time |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Horizon Site | Summary | Number of VDI Pools | vdiPoolCount | The number of VDI Pools in the Site at present |
| Number of RDS Desktop Pools | rdsDesktopPoolCount | The number of RDS Desktop Pools in the Site at present | ||
| Number of Application Pools | applicationPoolCount | The number of Application Pools in the Site at present | ||
| Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state | ||
| Number of Datastores | dsCount | Number of the vSphere Datastores | ||
| Capacity | capacity | Capacity of the Datastore in GB | ||
| Available Capacity | freeMemory | Free Capacity of the Datastore in GB | ||
| Number of Disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to | ||
| Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources | ||
| Number of Unhealthy Connection Servers | unHealthyConnectionServers | Number of Unhealthy Connection Servers under the Site | ||
| Number of Connection Servers | connectionServerCount | The number of Connection Servers in the Site at present. | ||
| Number of Pods | podCount | The number of pods in the Site at present. | ||
| Number of Farms | farmCount | Then number of RDS Farms in the Site at present | ||
| Number of RDS Hosts | rdsServerCount | The number of RDS Hosts in the Site at present. | ||
| Number of VDI Machines | vdiVirutalMachineCount | The number of VDI Virtual Machines in the Site at present | ||
| Performance | Datacenter KPI | datacenter | Average Datacenter KPI from all the RDS Hosts,VDI Sessions and Application Sessions from VDI Pools in the Site. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | |
| Network KPI | network | Average Network KPI from all the VDI, RDS and Application Sessions in the Site. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | ||
| Configuration | Number of Desktops in Bad State | noOfBadStateDesktops | Sum of all desktops in the pool that is not in ready or used state. Desktops that have problem should be 0 most of the time. | |
| Number of RDS Hosts in Bad State | noOfBadStateRDSHosts | Sum of all RDS Hosts in the farm that is not in ready or used state. RDS Hosts that have problem should be 0 most of the time | ||
| Utilization | Number of Users | totalUsers | The number of users who have at least one connected or disconnected session at present | |
| Number of Connected Users | connectedUsers | Number of users who have atleast one connected session at present. | ||
| Number of Disconnected Users | disconnectedUsers | Number of users who have only disconnected sessions. |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| HorizonWorld | UAG Servers | Datacenter KPI | datacenter | Average Datacenter KPI from all the Cloud Pod Federations. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) |
| Network KPI | network | Average Network KPI from all the Cloud Pod Federations. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | ||
| Number of UAG Servers with Red Network KPI | uagsWithRedNetworkKPI | The number of UAG Servers with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0. | ||
| Number of UAG Servers with Red Datacenter KPI | uagsWithRedDatacenterKPI | The number of UAG Servers with Datacenter KPI value falls within the red range (0% - 25%). Expect this number to be 0. | ||
| Number of UAG Servers with High CPU Ready | uagsWithHighCPUReady | Count of UAG Serves where CPU Ready > 95%. Expect this number to be minimal. Note that high performance typically requires high utilization | ||
| Number of UAG Servers with Low Available Memory | uagsWithLowAvailableMemory | Count of UAG Servers with Available Memory < 500 MB. Expect this number to be minimal for ideal performance. | ||
| Number of UAG Servers with High Disk Latency | uagsWithHighDiskLatency | Count of UAG Servers with Disk Latency > 15 ms. Expect this number to be minimal, especially on SSD-backed storage | ||
| Number of UAG Servers with High Protocol Packet Loss Transmit | uagsWithHighProtocolPacketLossTransmit | Count of UAG Servers with Horizon protocol (e.g. Blast, PCoIP) packet transmit loss > 1%. Expect this number to be low. In ideal state it is 0. | ||
| Number of UAG Servers with High Protocol Packet Loss Receive | uagsWithHighProtocolPacketLossReceive | Count of UAG Servers with Horizon protocol (e.g. Blast, PCoIP) packet receive loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Summary | Number of VDI Pools | vdiPoolCount | The number of VDI Pools in the inventory at present | |
| Number of RDS Desktop Pools | rdsDesktopPoolCount | The number of RDS Desktop Pools in the inventory at present | ||
| Number of Application Pools | applicationPoolCount | The number of Application Pools in the inventory at present | ||
| Number of Datastores | dsCount | Number of the vSphere Datastores | ||
| Capacity | capacity | Capacity of the Datastore in GB | ||
| Available Capacity | freeMemory | Free Capacity of the Datastore in GB | ||
| Number of Pods | podCount | The number of pods in the Horizon World at present | ||
| Number of RDS Farms | farmCount | The number of RDS Farms in the inventory at present | ||
| Number of RDS Hosts | rdsServerCount | The number of RDS Hosts in the inventory at present | ||
| Number of Connection Servers | connectionServerCount | The number of Horizon Connection Servers in the inventory. It counts any server regardless of operational status as long as it is registered with the Pod | ||
| Number of VDI Machines | vdiVirtualMachineCount | The number of VDI Virtual Machines in the inventory at present | ||
| Daily License Usage Count | dailyLicenseUsageCount | Capacity of the License being used by the User | ||
| Number of Orphaned Objects | noOfOrphanedObjects | Number of Orphaned Objects in the Horizon World | ||
| Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state | ||
| Number of Disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to | ||
| Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources | ||
| Number of Sites | siteCount | Number of Sites | ||
| Number of UAG Servers | gatewayCount | The present number of Unified Access Gateway in the inventory. This should match the documented architecture or plan. If not, either update the documentation or the actual deployment | ||
| Number of Clusters | clusterCount | Number of clusters in the World at present | ||
| Performance | Datacenter KPI | datacenter | Average Datacenter KPI from all the Cloud Pod Federations. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | |
| Network KPI | network | Average Network KPI from all the Cloud Pod Federations. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%) | ||
| Number of VDI Sessions with Red Network KPI | vdiSessionsWithRedNetworkKPI | The number of VDI Sessions with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of VDI Sessions with Red Datacenter KPI | vdiSessionsWithRedDatacenterKPI | The number of VDI Sessions with Datacenter KPI value falls within the red range (0% - 25%). Expect this number to be 0. | ||
| Number of VDI Pools with Red Network KPI | vdiPoolsWithRedNetworkKPI | The number of VDI Pools with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0. | ||
| Number of VDI Pools with Red Datacenter KPI | vdiPoolsWithRedDatacenterKPI | The number of VDI Pools with Datacenter KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of Pods with Red Datacenter KPI | podsWithRedDatacenterKPI | The number of Pod with Datacenter KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of Pods with Red Network KPI | podsWithRedNetworkKPI | The number of Pod with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0. | ||
| Number of Farms with Red Datacenter KPI | rdsFarmsWithRedDatacenterKPI | The number of RDS Farms with Datacenter KPI value falls within the red range (0% - 25%). Expect this number to be 0. | ||
| Number of Farms with Red Network KPI | rdsFarmsWithRedNetworkKPI | The number of RDS Farms with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0. | ||
| Number of Application Sessions with Red Network KPI | appSessionsWithRedNetworkKPI | The number of Application Sessions with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Number of RDS Sessions with Red Network KPI | rdsSessionsWithRedNetworkKPI | The number of VDI Pools with Network KPI value falls within the red range (0% - 25%). Expect this number to be 0 | ||
| Configuration | Number of Desktops in bad state | noOfBadStateDesktops | Sum of all desktops in the pool that is not in ready or used state. Desktops that have problem should be 0 most of the time | |
| Number of RDS Hosts in Bad State | noOfBadStateRDSHosts | Sum of all RDS Hosts in the farm that is not in ready or used state. RDS Hosts that have problem should be 0 most of the time | ||
| CPU | Number of VDI Sessions with High CPU Utilization | vdiSessionsWithHighCPUUtilization | Count of VDI Sessions where CPU Usage > 95%. Expect this number to be minimal. Note that high performance typically requires high utilization. | |
| Utilization | utilization | um of all VDI VMs and RDS Hosts CPU Utilization | ||
| Total Capacity | totalCapacity | Sum of all VDI VMs and RDS Hosts capacity that are part of Horizon | ||
| Disk | Number of VDI Sessions with High Disk Latency | vdiSessionsWithHighDiskLatency | Count of VDI Sessions with Disk Latency > 15 ms. Expect this number to be minimal, especially on SSD-backed storage | |
| Disk IOPS | diskIops | Sum of all VDI VMs and RDS Hosts I/O operations per second. It includes both Reads and Writes. Investigate if this becomes excessive | ||
| Read Disk IOPS | diskReadIops | Sum of all VDI VMs and RDS Hosts I/O operations per second. It includes only Read operations. Investigate if this becomes excessive | ||
| Write Disk IOPS | diskWriteIops | Sum of all VDI VMs and RDS Hosts I/O operations per second. It includes only Write operations. Investigate if this becomes excessive | ||
| Disk Throughput | diskThroughput | Sum of all VDI VMs and RDS Hosts disk throughput. It includes both Reads and Writes. Investigate if this becomes excessive. | ||
| Read Disk Throughput | dsikReadThroughput | Sum of all VDI VMs and RDS Hosts disk throughput. It includes only Reads. Investigate if this becomes excessive | ||
| Write Disk Throughput | diskWriteThroughput | Sum of all VDI VMs and RDS Hosts disk throughput. It includes only Writes. Investigate if this becomes excessive | ||
| Memory | Number of VDI Sessions with Low Available Memory | vdiSessionsWithLowAvailableMemory | Count of VDI Sessions with Available Memory < 500 MB. Expect this number to be minimal for ideal performance | |
| Utilization | utilization | Sum of all VDI VMs and RDS Hosts Memory Utilization | ||
| Total Capacity | totalCapacity | Sum of all VDI VMs and RDS Hosts capacity that are part of Horizon | ||
| Protocol | Number of Sessions with High Protocol Latency | sessionsWithHighProtocolLatency | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) latency > 180 ms. Expect this number to be low. In ideal state it is 0 | |
| Number of Sessions with High Protocol Packet Loss Transmit | sessionsWithHighProtocolPacketLossTransmit | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) packet transmit loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Number of Sessions with High Protocol Packet Loss Receive | sessionsWithHighProtocolPacketLossReceive | Count of Sessions (both RDS & VDI) with Horizon protocol (e.g. Blast, PCoIP) packet receive loss > 1%. Expect this number to be low. In ideal state it is 0 | ||
| Utilization | Number of Users | totalUsers | The number of users who have at least one connected or disconnected session at present. | |
| Number of Connected Users | connectedUsers | Number of users who have atleast one connected session at present. | ||
| Number of Inactive Users | inactiveUsers | Number of users who have connected to Horizon at least once but do not have connected or disconnected sessions at present | ||
| Number of Disconnected Users | disconnectedUsers | Number of users who have only disconnected sessions. | ||
| Number of Unique Users | uniqueUsers | The total number of unique users who are active and inactive. If a user has more than 1 session, it will be counted as 1 only. For Eg: A user with 1 VDI session and 5 Application session is counted as 1. |
| Resouce Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Connection Server | Summary | Connection Disparity | connectionDisparity | The % difference of Connections compared to Horizon Pod Average Connections per Connection Server. Disparity > 30% indicates the Connection Server has an abnormal workload and requires further analysis as to the cause |
| Connection | Number of Connections | actualConnection | The number of connections to this connection server | |
| Maximum Allowed Number of Connections | recommendedConnection | The maximum number of connections allowed for a connection server as per Horizon best practices | ||
| Connection Usage | connectionUsage | Connection Usage | ||
| Utilization | Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources. | |
| Horizon License | Licensed | licensed | Licensed | |
| Performance | CPU Usage Disparity | cpuUsageDisparity | Disparity of Connection Server CPU usage compared to the Pod Connection Server CPU usage Average | |
| Memory Usage Disparity | memoryUsageDisparity | Disparity of Connection Server memory usage compared to the Pod Connection Server memory usage Average | ||
| Telegraf | HTTP Health Check Status | httpStatusCode | Http Response Code | |
| HTTP Health Response Time | httpResponseTime | Http Health Response Time | ||
| CPU | Co-stop | costopAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Co-Stop among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time the VM is ready to run, but is unable due to co-scheduling constraints. VM with less vCPU have lower co-stop value. | |
| Ready | readyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Ready among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of CPU the VM is ready to run, but unable due to ESXi has no ready physical core to run it. High Ready value impacts VM performance | ||
| IO Wait | iowaitAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU IO Wait among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time VM CPU is waiting for IO. Formula is Wait - Idle - Swap Wait. High value indicates slow storage subsystem | ||
| Usage | usageAvg | CPU Usage divided by VM CPU Configuration in MHz | ||
| Run Queue | cpuqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Run Queue among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of ready threads queuing in the CPU. A number greater than 2 for prolong period indicates CPU core bottleneck. | ||
| Memory | Available Memory | freeMemoryAvg | Available memory on the guest OS computed as a sum of guest memory standby core, guest memory standby normal, guest memory standby reserve and guest memory free counters from the virtual machine. | |
| Page In Rate | pageInRatePerSecondAvg | The rate at which memory pages are paged in. An unexpected high value could indicate Guest OS does not have enough cache. This can cause performance problem for memory intensive application | ||
| Contention | contentionAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Memory Contention, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the amount of time, in percentage, the VM CPU is waiting for memory to be brought in. Keep this number below 1%. | ||
| Page Out Rate per second | pageOutRatePerSecondAvg | The rate at which memory pages are paged out. An unexpected high value could indicate Guest OS is under memory pressure. It's a capacity, not performance metric. | ||
| Disk | Disk Queue | diskqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Disk Queue Length, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of outstanding requests + IO currently in progress. | |
| Latency | totalLatencyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest latency among any of the virtual disks, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the weighted average of read latency and write latency | ||
| Worst vDisk Read Latency | peakvDiskReadLatency | Highest read latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Worst vDisk Write Latency | peakvDiskWriteLatency | Highest write latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Performance | Datacenter KPI | datacenter | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the datacenter segment, as opposed to the network. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Horizon User | Application | CPU | cpuUtilization | Average CPU utilization in percentage |
| Disk | diskUtilization | Average disk utilization in KB per second. | ||
| Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object. | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state. | ||
| Total Connected Session Time | connectedTime | The total time a user remains connected in this Pool | ||
| Logout Timestamp | logoffTime | The time at which the session on this Pool ends | ||
| RDS | Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object | |
| Logout Timestamp | logoffTime | The time the session ends. Used for historical reporting on the User object. | ||
| Total Connected Session Time | connectedTime | The total time a user remains connected out of all the sessions | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state. | ||
| VDI Desktop | Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object. | |
| Logout Timestamp | logoffTime | Time at which the last collection happened for events. | ||
| Idle Duration | idleDuraiton | The total time a user remains connected but in Idle state. | ||
| Total Connected Session Time | connectedTime | The total time a user remains connected in this Pool. | ||
| Is Oversized | isOverSized | Oversized = 1, Not oversized = 0. | ||
| Is Undersized | isUnderSized | Undersized =1, Not undersized =0. | ||
| Recommended CPU | recommendedCpu | CPU recommended by capacity model | ||
| Recommended Virtual CPUs | recommendedVcpu | Number of Virtual CPUs recommended by capacity model. | ||
| Recommended Memory | recommendedMemory | Amount of Memory recommended by capacity model. | ||
| Session | Worst time taken to load profile | worstLoadProfileTime | The longest time taken to load user profile among all the user sessions. | |
| Worst time taken to Logon | wortLogOnTime | The longest time taken to login to the desktop among all users in the VDI Pool. | ||
| Application/CPU | Co-stop | costopAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Co-Stop among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time the VM is ready to run, but is unable due to co-scheduling constraints. VM with less vCPU have lower co-stop value | |
| Ready | readyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Ready among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of CPU the VM is ready to run, but unable due to ESXi has no ready physical core to run it. High Ready value impacts VM performance | ||
| IO Wait | iowaitAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU IO Wait among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time VM CPU is waiting for IO. Formula is Wait - Idle - Swap Wait. High value indicates slow storage subsystem | ||
| Run Queue | cpuqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Run Queue among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of ready threads queuing in the CPU. A number greater than 2 for prolong period indicates CPU core bottleneck. | ||
| Usage | usageAvg | CPU Usage divided by VM CPU Configuration in MHz. | ||
| Worst Co-stop | worstCostopAvg | Worst Co-stop among the Sessions | ||
| Worst IO Wait | worstIOwaitAvg | Worst IO Wait among the Sessions. | ||
| Worst CPU Queue | worstCpuqueueAvg | Highest CPU Queue from all active VDI sessions. Keep this number below 3 queue per vCPU | ||
| Worst CPU Overlap | worstOverlapSummation | Highest CPU Overlap from all sessions | ||
| Worst vCPU Ready | worstPeakvcpuReady | The highest CPU Ready among the virtual CPU of the VM. Aim for this number below 2.5% for good user experience. | ||
| Application/Memory | Contention | contentionAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Memory Contention, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the amount of time, in percentage, the VM CPU is waiting for memory to be brought in. Keep this number below 1% | |
| Worst Page In Rate | worstPageInRatePerSecondAvg | The highest Page In Rate among the Sessions | ||
| Available Memory | freeMemoryAvg | Available memory on the guest OS computed as a sum of guest memory standby core, guest memory standby normal, guest memory standby reserve and guest memory free counters from the virtual machine. | ||
| Page In Rate | pageInRatePerSecondAvg | The rate at which memory pages are paged in. An unexpected high value could indicate Guest OS does not have enough cache. This can cause performance problem for memory intensive application. | ||
| Page Out Rate per second | pageOutRatePerSecondAvg | The rate at which memory pages are paged out. An unexpected high value could indicate Guest OS is under memory pressure. It's a capacity, not performance metric. | ||
| Worst Contention | worstContentionAvg | Worst Contention among the Sessions | ||
| Worst Available Memory | worstFreeMemoryAvg | Worst Available Memory among the Sessions | ||
| Application/Disk | Latency | totalLatencyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest latency among any of the virtual disks, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the weighted average of read latency and write latency | |
| Disk Queue | diskqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Disk Queue Length, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of outstanding requests + IO currently in progress. | ||
| Worst Outstanding IO | worstvDiskOIO | Highest disk outstanding IO from all active sessions. A user can have >1 sessions concurrently. Aim for this number to be below the underlying datastore disk queue length. | ||
| Worst vDisk Write Latency | peakvDiskWriteLatency | Highest VM virtual disk write latency from all active sessions. | ||
| Worst vDisk Read Latency | peakvDiskReadLatency | Highest VM virtual disk read latency from all active sessions. | ||
| Disk IOPS | diskIops | Sum of all VDI Pools I/O operations per second. It includes both Reads and Writes | ||
| Disk Throughput | diskThroughput | Sum of all VDI Pools disk throughput. It includes both Reads and Writes. | ||
| Worst Disk Queue | worstDiskqueueAvg | Worst Disk Queue among the Sessions. | ||
| Login Timestamp | loginTime | The actual time at which User logged on to a Horizon Session. | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state | ||
| Total Connected Session Time | connectedTime | The total time a user remains connected out of all the sessions. | ||
| VDI Desktop/CPU | Co-stop | costopAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Co-Stop among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time the VM is ready to run, but is unable due to co-scheduling constraints. VM with less vCPU have lower co-stop value | |
| Ready | readyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Ready among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of CPU the VM is ready to run, but unable due to ESXi has no ready physical core to run it. High Ready value impacts VM performance | ||
| IO Wait | iowaitAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU IO Wait among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time VM CPU is waiting for IO. Formula is Wait - Idle - Swap Wait. High value indicates slow storage subsystem | ||
| Run Queue | cpuqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Run Queue among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of ready threads queuing in the CPU. A number greater than 2 for prolong period indicates CPU core bottleneck. | ||
| Usage | usageAvg | CPU Usage divided by VM CPU Configuration in MHz. | ||
| Worst Co-stop | worstCostopAvg | Worst Co-stop among the Sessions | ||
| Worst IO Wait | worstIOwaitAvg | Worst IO Wait among the Sessions. | ||
| Worst CPU Queue | worstCpuqueueAvg | Highest CPU Queue from all active VDI sessions. Keep this number below 3 queue per vCPU | ||
| Worst CPU Overlap | worstOverlapSummation | Highest CPU Overlap from all sessions | ||
| Worst vCPU Ready | worstPeakvcpuReady | The highest CPU Ready among the virtual CPU of the VM. Aim for this number below 2.5% for good user experience. | ||
| Worst vCPU Usage | worstPeakvcpuUsage | The highest CPU utilization among the virtual CPU of the VM. An occasional high number is expected, but a prolonged high number coupled with CPU Run Queue and Context Switch can result in degraded user experience. | ||
| VDI Desktop/Memory | Contention | contentionAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Memory Contention, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the amount of time, in percentage, the VM CPU is waiting for memory to be brought in. Keep this number below 1% | |
| Page In Rate | pageInRatePerSecondAvg | The rate at which memory pages are paged in. An unexpected high value could indicate Guest OS does not have enough cache. This can cause performance problem for memory intensive application. | ||
| Available Memory | freeMemoryAvg | Available memory on the guest OS computed as a sum of guest memory standby core, guest memory standby normal, guest memory standby reserve and guest memory free counters from the virtual machine. | ||
| Worst Availabe Memory | worstFreeMemoryAvg | Worst Available Memory among the Sessions | ||
| Worst Page In Rate | worstPageInRatePerSecondAvg | The highest Page In Rate among the Sessions | ||
| Worst Contention | worstContentionAvg | Worst Contention among the Sessions | ||
| Page Out Rate per second | pageOutRatePerSecondAvg | The rate at which memory pages are paged out. An unexpected high value could indicate Guest OS is under memory pressure. It's a capacity, not performance metric. | ||
| VDI Desktop/Disk | Disk Queue | diskqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Disk Queue Length, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of outstanding requests + IO currently in progress. | |
| Latency | totalLatencyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest latency among any of the virtual disks, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the weighted average of read latency and write latency | ||
| Worst Disk Queue | worstDiskqueueAvg | Worst Disk Queue among the Sessions. | ||
| Worst Outstanding IO | worstvDiskOIO | Highest disk outstanding IO from all active sessions. A user can have >1 sessions concurrently. Aim for this number to be below the underlying datastore disk queue length. | ||
| Worst vDisk Read Latency | peakvDiskWriteLatency | Highest VM virtual disk write latency from all active sessions. | ||
| Worst vDisk Write Latency | peakvDiskReadLatency | Highest VM virtual disk read latency from all active sessions. | ||
| Disk IOPS | diskIops | Sum of all VDI Pools I/O operations per second. It includes both Reads and Writes | ||
| Disk Throughput | diskThroughput | Sum of all VDI Pools disk throughput. It includes both Reads and Writes. | ||
| VDI Desktop/Oversized | Memory | memory | Amount of additional Memory required on this pool for this User. | |
| Virtual CPUs | vCpus | Number of additional vCPUs required on this pool for this User | ||
| VDI Desktop/Undersized | Memory | memory | Amount of additional Memory than needed on this pool by this User. | |
| Virtual CPUs | vCpus | Number of additional vCPUs than allocated used on this pool by this User. | ||
| Protocol | Latency | latency | The time taken for Horizon protocol packets to reach destination. While related, it is not the same with protocol (e.g. TCP/IP) latency. It's a normalized average of transmit and receive packets. | |
| Frame Rate | frameRate | The frequency of screen frame change per second. Higher refresh rates results in smoother visual but it requires more bandwidth. | ||
| packet Loss Transmit | packetLossTransmit | The percentage of Horizon protocol packets that did not reach Horizon Client or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Packet Loss Receive | packetLossReceive | The percentage of Horizon protocol packets that did not reach Horizon Agent or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Worst Latency | worstLatency | Worst Latency from all active sessions. A user can have >1 sessions concurrently. | ||
| Worst Frame Rate | worstFrameRate | Worst Frame Rate from all active sessions. A user can have >1 sessions concurrently. | ||
| Worst Packet Loss Transmit | worstPacketLossTransmit | Highest transmitted packet loss from all active sessions. A user can have >1 sessions concurrently. Keep this number below 1%. | ||
| Worst Packet Loss Receive | worstPacketLossReceive | Worst Packet Loss Receive from all active sessions. A user can have >1 sessions concurrently. | ||
| RDS | CPU Usage | overallCpu | The session CPU consumption relative to the RDS Host size. This is not the total RDS Host usage. | |
| RDS Desktop Pool Names | desktopPoolName | RDS Desktop Pools that the user currently use | ||
| RDS Server MOIDs | rdsHostmoid | RDS Server MOIDs | ||
| RDS Server DNS Names | rdsHostDnsName | RDS Server DNS Names. | ||
| Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object. | ||
| Total Connected Session Time | connectedTime | The total time a user remains connected in this Pool. | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state. | ||
| Logout Timestamp | logoffTime | The time the session ends. Used for historical reporting on the User object. | ||
| Performance | Worst KPI | worstKPI | Worst value of VDI Session and RDS Session and Application Session, where VDI Session = Average (Datacenter KPI + Network KPI). | |
| Worst Datacenter KPI | worstDatacenter | Lowest Datacenter KPI from all active sessions. A user can have >1 sessions concurrently. Aim for this number to be in the green range (75% - 100%) | ||
| Worst Network KPI | worstNetwork | Lowest Network KPI from all active sessions. A user can have >1 sessions concurrently. Aim for this number to be in the green range (75% - 100%) | ||
| Summary | Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state | |
| Number of Disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to. | ||
| Number of Sessions | noOfSessions | Count of Sessions connected + disconnected. Used for understanding the number of sessions currently reserving or actively consuming Horizon resources. |
| Resource Name | MetricGroup | Metric Key | Metric | Description |
|---|---|---|---|---|
| VDI Desktop Session | Session | Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object |
| Total Connected Sesison Time | connectedTime | The total time a user remains connected across the pool or farm | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state | ||
| Time taken to Logon | logonDuration | The longest time taken to login to the desktop/application among all the sessions. | ||
| Time taken to load Profile | userProfileLoadDuration | The longest time taken to load user profile among all the sessions | ||
| Protocol | Latency | latency | The time taken for Horizon protocol packets to reach destination. While related, it is not the same with protocol (e.g. TCP/IP) latency. It's a normalized average of transmit and receive packets. | |
| Frame Rate | frameRate | The frequency of screen frame change per second. Higher refresh rates results in smoother visual but it requires more bandwidth. | ||
| Packet Loss Receive | packetLossReceive | The percentage of Horizon protocol packets that did not reach Horizon Agent or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Packet Loss Trasmit | packetLossTransmit | The percentage of Horizon protocol packets that did not reach Horizon Client or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Transmit Bandwidth Utilization | transmittedBandwidth | The average network bandwidth transmit utilization in the collection period. Investigate if there is prolonged excessive usage. | ||
| Summary | Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state. | |
| Number of Disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to | ||
| Is Oversized | isOverSized | Oversized = 1, Not oversized = 0 | ||
| Is Undersized | isUnderSized | Undersized =1, Not undersized =0 | ||
| Number of PCOIP Sessions | pcoipSessions | Number of PCOIP sessions | ||
| Number of BLAST Sessions | blastSessions | Number of BLAST sessions | ||
| Oversized | Memory | memory | Amount of Additional Memory used than Allocated. | |
| Virtual CPUs | vCpus | Number of additional vCPUs used than Allocated | ||
| Undersized | Memory | memory | Amount of Additional Memory Required | |
| Virtual CPUs | vCpus | Number of Additional vCPUs Required | ||
| VM Names | vmName | Name of the Virtual Machine where the session is created. | ||
| User Name | userName | Name of logged in User. | ||
| CPU | Co-stop | costopAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Co-Stop among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time the VM is ready to run, but is unable due to co-scheduling constraints. VM with less vCPU have lower co-stop value. | |
| Ready | readyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Ready among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of CPU the VM is ready to run, but unable due to ESXi has no ready physical core to run it. High Ready value impacts VM performance. | ||
| IO Wait | iowaitAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU IO Wait among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time VM CPU is waiting for IO. Formula is Wait - Idle - Swap Wait. High value indicates slow storage subsystem. | ||
| Run Queue | cpuqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Run Queue among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of ready threads queuing in the CPU. A number greater than 2 for prolong period indicates CPU core bottleneck | ||
| Usage | usageAvg | Amount of actively used virtual CPU. This is the host's view of the CPU usage not the guest operating system view. | ||
| Context Switch | contextSwapRateLatest | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Context Switch Rate, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the rate at which Operating System processes running in a CPU are loaded and unloaded. As context switch is an overhead cost, a high number will impact the application performance. The number varies per application so profile your environment to baseline the expected range. | ||
| Overlap | overlapSummation | Percentage of time where the VM was interrupted by hypervisor as it needs to perform system services on behalf of that VM or other VMs. | ||
| Peak vCPU Ready | peakvcpuReady | The highest CPU Ready among the virtual CPU of the VM. Aim for this number below 2.5% for good user experience. | ||
| Peak vCPU Usage | peakvcpuUsage | The highest CPU utilization among the virtual CPU of the VM. An occasional high number is expected, but a prolonged high number coupled with CPU Run Queue and Context Switch can result in degraded user experience | ||
| Utilization | utilization | Amount of actively used virtual CPU. This is the host's view of the CPU usage not the guest operating system view | ||
| Configured Capacity | totalCapacity | Memory resources allocated to the Virtual Machine. | ||
| Virtual CPUs | vCpus | Number of vCPUs | ||
| Recommended Virtual CPUs | recommendedVcpu | Number of Virtual CPUs recommended by capacity model | ||
| Memory | Contention | contentionAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Memory Contention, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the amount of time, in percentage, the VM CPU is waiting for memory to be brought in. Keep this number below 1%. | |
| Available Memory | freeMemoryAvg | Available memory on the guest OS computed as a sum of guest memory standby core, guest memory standby normal, guest memory standby reserve and guest memory free counters from the virtual machine. | ||
| Page In Rate | pageInRatePerSecondAvg | The rate at which memory pages are paged in. An unexpected high value could indicate Guest OS does not have enough cache. This can cause performance problem for memory intensive application | ||
| Utilization | utilization | Amount of memory utilized by the Virtual Machine. Reflects the guest OS memory required (for certain vSphere and VMTools versions) or Virtual Machine consumption | ||
| Configured Capacity | totalCapacity | Memory resources allocated to the Virtual Machine. | ||
| Page Out Rate per second | pageOutRatePerSecondAvg | Number of Pages per Second | ||
| Disk | Latency | totalLatencyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest latency among any of the virtual disks, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the weighted average of read latency and write latency. | |
| Disk Queue | diskqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Disk Queue Length, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of outstanding requests + IO currently in progress. | ||
| Outstanding IO | vDiskOIO | Amount of disk Input or Output commands waiting in the queue to be executed. High IO, coupled with high latency, impacts performance. This number should be below the queue depth of the OS. | ||
| Worst vDisk Read Latency | peakvDiskReadLatency | Highest read latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Worst vDisk Write Latency | peakvDiskWriteLatency | Highest write latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Disk IOPS | diskIops | Number of read/write operations per second. This is averaged over the reporting period. | ||
| Read Disk IOPS | diskReadIops | Number of read operations per second. This is averaged over the reporting period. | ||
| Write Disk IOPS | diskWriteIops | Number of write operations per second. This is averaged over the reporting period. | ||
| Disk Throughput | diskThroughput | Amount of data read from/written to storage in a second. This is averaged over the reporting period. | ||
| Read Disk Throughput | diskReadThroughput | Amount of data read from storage in a second. This is averaged over the reporting period. | ||
| Write Disk Throughput | diskWriteThroughput | Amount of data written to storage in a second. This is averaged over the reporting period. | ||
| Performance | Datacenter KPI | datacenter | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the datacenter segment, as opposed to the network. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). | |
| Network KPI | network | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the network segment of session, as opposed to the datacenter. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). | ||
| Performance KPI | averageKPI | Average of Datacenter and Network KPI Metrics. |
| Resource Name | MetricGroup | Metric Key | Metric | Description |
|---|---|---|---|---|
| RDS Desktop Session | Session | Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object. |
| Total Connected Session Time | connectedTime | The total time user remains connected. | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state. | ||
| Time taken to Logon | logonDuration | The total time user remains connected across pool or farm. | ||
| Time taken to load Profile | userProfileLoadDuration | The longest time taken to load user profile among all the sessions. | ||
| Protocol | Latency | latency | The time taken for Horizon protocol packets to reach destination. While related, it is not the same with protocol (e.g. TCP/IP) latency. It's a normalized average of transmit and receive packets. | |
| Frame Rate | frameRate | The frequency of screen frame change per second. Higher refresh rates results in smoother visual but it requires more bandwidth. | ||
| Packet Loss Receive | packetLossReceive | The percentage of Horizon protocol packets that did not reach Horizon Agent or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Packet Loss Trasmit | packetLossTransmit | The percentage of Horizon protocol packets that did not reach Horizon Client or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state. | ||
| Number of Disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to | ||
| Transmit Bandwidth Utilization | transmittedBandwidth | The average network bandwidth transmit utilization in the collection period. Investigate if there is prolonged excessive usage. | ||
| Number of BLAST Sessions | blastSessions | Number of BLAST sessions | ||
| Number of PCOIP Sessions | pcoipSessions | Number of PCOIP sessions | ||
| User Name | userName | Name of logged in User. | ||
| VM Names | vmName | Name of the Virtual Machine where the session is created. | ||
| RDS Application | CPU Usage | overallCpu | The session CPU consumption relative to the RDS Host size. This is not the total RDS Host usage. | |
| Disk IOPS | avgDiskIops | The average of read + write operations per second. Investigate if this becomes excessive. This is a value reported by Guest OS, not VM, as it is a session. | ||
| Overall Memory | overallMemory | Memory Utilization | ||
| Performance | Network KPI | network | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the network segment of session, as opposed to the datacenter. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). |
| Resource Name | MetricGroup | Metric Key | Metric | Description |
|---|---|---|---|---|
| VDI Application Session | Session | Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object |
| Packet Loss Receive | packetLossReceive | The percentage of Horizon protocol packets that did not reach Horizon Agent or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Frame Rate | frameRate | The frequency of screen frame change per second. Higher refresh rates results in smoother visual but it requires more bandwidth. | ||
| Transmit Bandwidth Utilization | transmittedBandwidth | The average network bandwidth transmit utilization in the collection period. Investigate if there is prolonged excessive usage. | ||
| Packet Loss Trasmit | packetLossTransmit | The percentage of Horizon protocol packets that did not reach Horizon Client or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Total Connected Session Time | connectedTime | The total time user remains connected. | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state | ||
| Time taken to Logon | logonDuration | The longest time taken to login to the desktop/application among all the sessions. | ||
| Time taken to load Profile | userLoadProfileDuration | The longest time taken to load user profile among all the sessions | ||
| Protocol | Latency | latency | The time taken for Horizon protocol packets to reach destination. While related, it is not the same with protocol (e.g. TCP/IP) latency. It's a normalized average of transmit and receive packets. | |
| Number of Disconnected Sessions | disconnectedSessions | The number of sessions in a pool where the user has established a desktop session, but is not currently connected. These desktops will not be available for other users to connect to | ||
| Summary | Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state. | |
| Number of BLAST Sessions | blastSessions | Number of BLAST sessions | ||
| VM Names | vmName | Name of the Virtual Machine where the session is created. | ||
| Number of PCOIP Sessions | pcoipSessions | Number of PCOIP sessions | ||
| Application | CPU Utilization | cpu | Average CPU utilization of all applications in the session in percentage.Metric is calculated using help desk APIs | |
| Cpu | Utilization | utilization | Amount of actively used virtual CPU. This is the host's view of the CPU usage not the guest operating system view | |
| Configured Capacity | totalCapacity | Configured Capacity in GHz, based on nominal (static) frequency of the CPU. | ||
| Co-stop | costopAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Co-Stop among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time the VM is ready to run, but is unable due to co-scheduling constraints. VM with less vCPU have lower co-stop value. | ||
| Ready | readyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Ready among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of CPU the VM is ready to run, but unable due to ESXi has no ready physical core to run it. High Ready value impacts VM performance. | ||
| IO Wait | iowaitAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU IO Wait among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time VM CPU is waiting for IO. Formula is Wait - Idle - Swap Wait. High value indicates slow storage subsystem. | ||
| Run Queue | cpuqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Run Queue among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of ready threads queuing in the CPU. A number greater than 2 for prolong period indicates CPU core bottleneck | ||
| Usage | usageAvg | Amount of actively used virtual CPU. This is the host's view of the CPU usage not the guest operating system view. | ||
| Context Switch | contextSwapRateLatest | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Context Switch Rate, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the rate at which Operating System processes running in a CPU are loaded and unloaded. As context switch is an overhead cost, a high number will impact the application performance. The number varies per application so profile your environment to baseline the expected range. | ||
| Overlap | overlapSummation | Percentage of time where the VM was interrupted by hypervisor as it needs to perform system services on behalf of that VM or other VMs. | ||
| Peak vCPU Ready | peakvcpuReady | The highest CPU Ready among the virtual CPU of the VM. Aim for this number below 2.5% for good user experience. | ||
| Peak vCPU Usage | peakvcpuUsage | The highest CPU utilization among the virtual CPU of the VM. An occasional high number is expected, but a prolonged high number coupled with CPU Run Queue and Context Switch can result in degraded user experience | ||
| Memory | Utilization | utilization | Amount of memory utilized by the Virtual Machine. Reflects the guest OS memory required (for certain vSphere and VMTools versions) or Virtual Machine consumption | |
| Configured Capacity | totalCapacity | Memory resources allocated to the Virtual Machine. | ||
| Contention | contentionAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Memory Contention, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the amount of time, in percentage, the VM CPU is waiting for memory to be brought in. Keep this number below 1%. | ||
| Available Memory | freeMemoryAvg | Available memory on the guest OS computed as a sum of guest memory standby core, guest memory standby normal, guest memory standby reserve and guest memory free counters from the virtual machine. | ||
| Page In Rate | pageInRatePerSecondAvg | The rate at which memory pages are paged in. An unexpected high value could indicate Guest OS does not have enough cache. This can cause performance problem for memory intensive application | ||
| Page Out Rate per second | pageOutRatePerSecondAvg | Number of Pages per second | ||
| Disk | Disk IOPS | diskIops | Number of read/write operations per second. This is averaged over the reporting period. | |
| Read Disk IOPS | diskReadIops | Number of read operations per second. This is averaged over the reporting period. | ||
| Write Disk IOPS | diskWriteIops | Number of write operations per second. This is averaged over the reporting period. | ||
| Disk Throughput | diskThroughput | Amount of data read from/written to storage in a second. This is averaged over the reporting period. | ||
| Read Disk Throughput | diskReadThroughput | Amount of data read from storage in a second. This is averaged over the reporting period. | ||
| Write Disk Throughput | diskWriteThroughput | Amount of data written to storage in a second. This is averaged over the reporting period. | ||
| Latency | totalLatencyAvg | The time taken for Horizon protocol packets to reach destination. While related, it is not the same with protocol (e.g. TCP/IP) latency. It's a normalized average of transmit and receive packets. | ||
| Disk Queue | diskqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Disk Queue Length, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of outstanding requests + IO currently in progress. | ||
| Outstanding IO | vDiskOIO | Amount of disk Input or Output commands waiting in the queue to be executed. High IO, coupled with high latency, impacts performance. This number should be below the queue depth of the OS. | ||
| Worst vDisk Read Latency | peakvDiskReadLatency | Highest read latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Worst vDisk Write Latency | peakvDiskWriteLatency | Highest write latency among any of the VM virtual disks. Keep this number below 10 ms. | ||
| Performance | Network KPI | network | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the network segment of session, as opposed to the datacenter. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). | |
| Datacenter KPI | datacenter | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the datacenter segment, as opposed to the network. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). | ||
| Performance KPI | averageKPI | Average of Datacenter and Network KPI Metrics. |
| Resource Name | MetricGroup | Metric Key | Metric | Description |
|---|---|---|---|---|
| RDS Application Session | Session | Login Timestamp | loginTime | The actual time the session starts. Used for historical reporting on the User object |
| Total Connected Session Time | connectedTime | The total time user remains connected. | ||
| Idle Duration | idleDuration | The total time a user remains connected but in Idle state | ||
| Time taken to Logon | logOnDuration | The longest time taken to login to the desktop/application among all the sessions. | ||
| Time taken to load Profile | userProfileLoadDuration | The longest time taken to load user profile among all the sessions | ||
| Protocol | Latency | latency | The time taken for Horizon protocol packets to reach destination. While related, it is not the same with protocol (e.g. TCP/IP) latency. It's a normalized average of transmit and receive packets. | |
| Frame Rate | frameRate | The frequency of screen frame change per second. Higher refresh rates results in smoother visual but it requires more bandwidth. | ||
| Packet Loss Receive | packetLossTransmit | The percentage of Horizon protocol packets that did not reach Horizon Agent or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Packet Loss Trasmit | packetLossReceive | The percentage of Horizon protocol packets that did not reach Horizon Client or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Application | CPU Utilization | cpu | Average CPU utilization of all applications in the session in percentage.Metric is calculated using help desk APIs | |
| Summary | Number of Connected Sessions | connectedSessions | Number of sessions at present that are in connected state. | |
| Number of BLAST Sessions | blastSessions | Number of BLAST sessions | ||
| VM Names | vmName | Name of the Virtual Machine where the session is created. | ||
| User Name | userName | Name of logged in User. | ||
| Number of PCOIP Sessions | pcoipSessions | Number of PCOIP sessions |
| Resource Name | Metric Group | Metric Key | Metric | Description |
|---|---|---|---|---|
| Unified Access Gateway | Summary | Number of Connected Sessions | connectedSessions | Number of Connected Sessions passed through UAG |
| Number of Sessions Utilized | sessionsUtilized | % is based on Number of sessions out of 2000 max value. | ||
| Number of PCOIP Sessions | pcoipSessions | Number of PCOIP sessions | ||
| Number of BLAST Sessions | blastSessions | Number of BLAST Sessions | ||
| Telegraf | HTTP Health Check Status | httpStatusCode | Http Response Code | |
| HTTP Health Response Time | httpResponseTime | Http Health Response Time | ||
| Protocol | Latency | latency | The time taken for Horizon protocol packets to reach destination. While related, it is not the same with protocol (e.g. TCP/IP) latency. It's a normalized average of transmit and receive packets. | |
| Frame Rate | frameRate | The frequency of screen frame change per second. Higher refresh rates results in smoother visual but it requires more bandwidth. | ||
| Packet Loss Transmit | packetLossTransmit | The percentage of Horizon protocol packets that did not reach Horizon Client or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Packet Loss Receive | packetLossReceive | The percentage of Horizon protocol packets that did not reach Horizon Agent or was dropped due to out of order arrival. This number directly impacts user experience. | ||
| Worst Latency | worstLatency | Worst Latency from all active sessions. A user can have >1 sessions concurrently | ||
| Worst Frame Rate | worstFrameRate | Worst Frame Rate from all active sessions. A user can have >1 sessions concurrently. | ||
| Worst Packet Loss Transmit | worstPacketLossTransmit | Highest transmitted packet loss from all active sessions. A user can have >1 sessions concurrently. Keep this number below 1% | ||
| Worst Packet Loss Receive | worstPacketLossReceive | Worst Packet Loss Receive from all active sessions. A user can have >1 sessions concurrently. | ||
| Number of Sessions with High Protocol Packet Transmit Loss | sessionsWithHighProtocolPacketLossTransmit | Count of Sessions with Horizon protocol (e.g. Blast, PCoIP) packet transmit loss > 1%. Expect this number to be low. In ideal state it is 0. | ||
| Number of Sessions with High Protocol Packet Receive Loss | sessionsWithHighProtocolPacketLossReceive | Count of Sessions with Horizon protocol (e.g. Blast, PCoIP) packet receive loss > 1%. Expect this number to be low. In ideal state it is 0. | ||
| Number of Sessions with High Protocol Latency | sessionsWithHighProtocolLatency | Count of Sessions with Horizon protocol (e.g. Blast, PCoIP) latency > 180 ms. Expect this number to be low. In ideal state it is 0. | ||
| CPU | Co-stop | costopAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Co-Stop among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time the VM is ready to run, but is unable due to co-scheduling constraints. VM with less vCPU have lower co-stop value. | |
| Ready | readyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Ready among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of CPU the VM is ready to run, but unable due to ESXi has no ready physical core to run it. High Ready value impacts VM performance. | ||
| IO Wait | iowaitAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU IO Wait among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the percentage of time VM CPU is waiting for IO. Formula is Wait - Idle - Swap Wait. High value indicates slow storage subsystem. | ||
| Usage | usageAvg | CPU Usage divided by VM CPU Configuration in MHz. | ||
| Run Queue | cpuqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest CPU Run Queue among any of the vCPUs, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of ready threads queuing in the CPU. A number greater than 2 for prolong period indicates CPU core bottleneck | ||
| Memory | Available Memory | freeMemoryAvg | Available memory on the guest OS computed as a sum of guest memory standby core, guest memory standby normal, guest memory standby reserve and guest memory free counters from the virtual machine. | |
| Page In Rate | pageInRatePerSecondAvg | The rate at which memory pages are paged in. An unexpected high value could indicate Guest OS does not have enough cache. This can cause performance problem for memory intensive application. | ||
| Contention | contentionAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Memory Contention, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the amount of time, in percentage, the VM CPU is waiting for memory to be brought in. Keep this number below 1%. | ||
| Page Out Rate per second | pageOutRatePerSecondAvg | The rate at which memory pages are paged out. An unexpected high value could indicate Guest OS is under memory pressure. It's a capacity, not performance metric. | ||
| Utilization | utilization | Amount of memory utilized by the Virtual Machine. Reflects the guest OS memory required (for certain vSphere and VMTools versions) or Virtual Machine consumption. | ||
| Configured Capacity | totalCapacity | Memory resources allocated to the Virtual Machine. | ||
| Usage | usageAvg | Memory currently in use as a percentage of total available memory | ||
| Disk | Disk Queue | diskqueueAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest Disk Queue Length, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the number of outstanding requests + IO currently in progress. | |
| Latency | totalLatencyAvg | If you have enabled the peak metric collection for vCenter VMs, this metric is set to the highest latency among any of the virtual disks, measured as peak of any 20-second average during the collection interval. If the peak metrics are not enabled, this metric is set to the weighted average of read latency and write latency. | ||
| Read Latency | readLatencyAvg | Average amount of time for a read operation by the storage adapter. | ||
| Write Latency | writeLatencyAvg | Average amount of time for a write operation by the storage adapter. | ||
| Performance | Datacenter KPI | datacenter | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the network segment of session, as opposed to the datacenter. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). | |
| Network KPI | Network | The Key Performance Indicator, aggregating all the performance metrics into 1 metric. This focuses on the datacenter segment, as opposed to the network. Value is 0 - 100% and divided into 4 levels (Green 75 - 100%, Yellow 50 - 75%, Orange 25 - 50%, Red 0 - 25%). | ||
| Performance KPI | averageKPI | Average of Datacenter and Network KPI Metrics. | ||
| Network | Total Transmitted Packets Dropped | totalDroppedTxPackets | Number of transmitted packets dropped based on summation of 20 sec samples during the monitoring interval | |
| Transmitted Packets Dropped | droppedTxPackets | Number of outgoing packets dropped based on average of 20 sec samples during the monitoring interval. Investigate if the number is not 0. |
