If you are deploying VMware Telco Cloud Service Assurance in the High Availability (HA) mode, ensure that your system meets the following requirements and the deployment scaling requirements.
Footprint Specification
The following tables specify the requirements for deploying different footprints. The number of VMs required for each type of cluster is provided. This is particularly important when HA is required. The following tables provides the number of virtual CPUs, main memory (RAM), and the total disk size for each virtual machine.
| Footprint Size | Number of VMs | vCPU Per VM | RAM Per VM (GBs) | Role |
|---|---|---|---|---|
| 25 K | 3 | 2 | 8 | Control Plane Node |
| 2 | 2 | 8 | Worker Node | |
| 50 K | 3 | 4 | 16 | Control Plane Node |
| 2 | 4 | 16 | Worker Node | |
| 100 K | 3 | 4 | 16 | Control Plane Node |
| 2 | 4 | 16 | Worker Node |
Note: The table shows the
VMware Tanzu Kubernetes Grid management cluster sizing for deployments when a dedicated
VMware Tanzu Kubernetes Grid management cluster is used for the
VMware Tanzu Kubernetes Grid workload cluster in
VMware Telco Cloud Service Assurance. To size deployments when multiple workload clusters are managed by a single management cluster, see
VMware Tanzu Kubernetes Grid Documentation.
| Footprint Size | Number of VMs | vCPU Per VM | RAM Per VM (GBs) | Local Disk Per VM (GBs) | Total Persistent Volume Storage (TBs) for 1W Retention Interval | Role |
|---|---|---|---|---|---|---|
| 25 K | 3 | 2 | 8 | 50 | NA | Control Plane Node |
| 9 | 16 | 64 | 200 | 13.5 | Worker Node | |
| 50 K | 3 | 4 | 16 | 50 | NA | Control Plane Node |
| 14 | 16 | 64 | 200 | 19 | Worker Node | |
| 100 K | 3 | 4 | 16 | 50 | NA | Control Plane Node |
| 20 | 16 | 64 | 200 | 35.5 | Worker Node |
Note: For VMware Tanzu Kubernetes management and workload cluster, it is recommended to enable vSphere HA for VMWare cluster.
| Footprint Size | Number of VMs | vCPU Per VM | RAM Per VM (GBs) | Local Disk Per VM (GBs) | Total Persistent Volume Storage (TBs) for 1W Retention Interval |
|---|---|---|---|---|---|
| 25 K | 9 | 16 | 64 | 200 | 13.5 |
| 50 K | 14 | 16 | 64 | 200 | 19 |
| 100 K | 20 | 16 | 64 | 200 | 35.5 |
Note: By default in AKS, first three worker nodes can also be the control plane nodes.
For 25 K, 50 K, and 100 K footprints, the recommended AKS VM size template is Standard_D16s_v3.
Raw Metric Retention Interval
The following table provides total persistent volume storage in Terra Bytes (TBs) required for 25 K, 50 K, and 100 K footprints with raw metrics based on the retention period, which varies from 1 week through 7 weeks.
Note: The default raw metrics retention interval is 1w and is configurable. If you decide to change the retention interval period, the persistent volume storage must change accordingly.
| Footprint | Raw Metric Retention Interval in Weeks | ||||||
|---|---|---|---|---|---|---|---|
| PV Storage (TBs) for 1W | PV Storage (TBs) for 2W | PV Storage (TBs) for 3W | PV Storage (TBs) for 4W | PV Storage (TBs) for 5W | PV Storage (TBs) for 6W | PV Storage (TBs) for 7W | |
| 25 K | 13.5 | 13.5 | 13.5 | 19.2 | 19.2 | 19.2 | 19.2 |
| 50 K | 19 | 24.2 | 24.2 | 29.6 | 29.6 | 36 | 36 |
| 100 K | 35.5 | 40.5 | 45.8 | 51.5 | 56.8 | 62.5 | 69 |
For example, if the raw metrics for the retention period is 4 weeks for a 50 K footprint, make sure that 29.6 TB persistent volume storage is provisioned. During the deployment time, specify the retention period as 4 weeks.
Performance and Scalability for Different Deployments
The following table provides sample managed capacity for each footprint.
| Footprint | Small (HA) 25 K | Small-Medium (HA) 50 K | Medium (HA) 100 K |
|---|---|---|---|
| Number of devices | 25 K | 50 K | 100 K |
| Number of unique events or notifications per day | 25 K | 50 K | 100 K |
| Number of metrics per five minutes | 10 million | 20 million | 40 million |
| Number of routers or switches | 15 K | 30 K | 60 K |
| Managed P and I | 300 K | 600 K | 1.2 million |
| Number of hosts | 2 K | 4 K | 8 K |
| Number of VMs | 5 K | 10 K | 20 K |
| Number of CNFs | 10 K | 10 K | 10 K |
| Number of pods | 10 K | 10 K | 10 K |
| Total number of events=number of devices*4+external events | 105 K | 205 K | 500 K |
| Number of raw metrics from Domain Manager metric collector per five minutes polling interval | 9 million | 14 million | 29 million |
| Number of Cisco ACI control cluster supported | 10 | 10 | 10 |
| Number of IPSLA routers discovered per collector | 100 | 100 | 100 |
| Number of Analytics jobs supported | 10 | 10 | 10 |
| Number of Alarming jobs supported | 10 | 10 | 10 |
| Total number of alerts supported per minute | 100 | 100 | 100 |
| Kafka to Kafka collector metrics per five minutes polling interval | 1 million | 6 million | 11 million |
| Total number of concurrent APIs | 100 | 100 | 100 |
| Number of concurrent users | 10 | 10 | 10 |
| Total number of users | 200 | 200 | 200 |
| Maximum number of events from VMware vROps to VMware Telco Cloud Service Assurance per five minutes polling interval | 6 K | 6 K | 6 K |
| Number of notifications processed per second | 350 | 450 | 450 |
| Data synchronization of Topology in VMware Telco Cloud Service Assurance UI | 6 minutes | 8 minutes | 10 minutes |
| Number of metrics that can be exported to external Kafka per five minutes polling interval | 10 Million | 20 Million | 40 Million |
| Bandwidth utilization for storage traffic | 33 Mbps | 65 Mbps | 135 Mbps |
| Total Disk IOPS (Read + Write) | 1000 | 2000 | 4000 |
| Footprint | Small (HA) 25 K | Small-Medium (HA) 50 K | Medium (HA) 100 K |
|---|---|---|---|
| Native traffic flow metrics | 1.5 K | 2.5 K | 5 K |
Note: For native traffic flow, scale support mentioned in the table considers that no other source of metric data is flowing to
VMware Telco Cloud Service Assurance. If other source of metric data is flowing into
VMware Telco Cloud Service Assurance, then reduce the ratio of traffic flow data.
