VMware vRealize Operations Manager Appliance 6.4 | 15 November 2016 | Build 4635874
VMware vRealize Operations Manager 6.4 for Windows and Linux | 15 November 2016 | Build 4635875
VMware vRealize Operations Manager 6.4 PAK Updates | 15 November 2016 | Build 4635873
VMware vRealize Operations Manager 6.4 Endpoint Operations Agents | 15 November 2016 | Build 4536552
Updated on: 28 Mar 2017
Check frequently for additions and updates to these release notes.
What's in the Release Notes?
The release notes cover the following topics:
- What's New?
- System Requirements
- Installing and Updating vRealize Operations Manager
- Resolved Issues from Previous Releases
- Known Issues in This Release
- Known Issues From Previous Releases
New Patches to Address Security Advisory
- On March 13, 2017, a critical severity privilege escalation was disclosed. The details of this vulnerability are documented in VMSA-2017-0004. To address this issue, install the public patch documented in KB 2149591.
vRealize Operations Manager 6.4 focuses on enhancing product usability and stability and supporting vSphere 6.5 environments.
- Pinpoints and helps you clean up alerts in your environments:
- Groups alerts by the definition to identify noisiest alerts.
- Allows single-click disabling of alerts across the system or for a given policy.
- Reflects business context based on vSphere VM folders:
- vSphere VM folders are now incorporated into the navigation tree for quick searching.
- Thirteen new installed dashboards to display status, identify problems, and at-a-glance share data with your peers:
- A getting started dashboard to introduce you to using our dashboards
- Environment and capacity overview dashboards to get a summary of your environments.
- VM troubleshooting dashboard that helps you diagnose problems in a VM and start solving them.
- Infrastructure capacity and performance dashboards to view status and see problems across your datacenter.
- VM and infrastructure configuration dashboards to highlight inconsistencies and violations of VMware best practices in your environment.
- Enhanced All Metrics tab to ease troubleshooting:
- Set of key KPIs per object and resource for a quick start of your investigation process.
- Ability to correlate properties with metrics
- Support for vSphere 6.5
- Summary pages for key vSphere objects:
- Deeper information about the objects. For VMs, vRealize Operations Manager provides OS version, IP address, tags, and other details.
- Show known problems on an object or on its children.
- Display key performance indicators for these objects.
- Highlight problems with related objects with the same infrastructure, such as neighbors of a VM or children of a host.
- Predictive DRS (pDRS) enables vRealize Operations to provide long-term forecast metrics to augment the short-term placement criteria of the vSphere Distributed Resource Scheduler (DRS). This capability only works within vSphere version 6.5 or above. Note that your software warranty and support policy only applies when operating pDRS in a cluster containing 4,000 or fewer VMs (see section 5.3.1 of the VMware Product Guide.
For more information on these features, see the vRealize Operations Manager 6.4 VMware product page
Review this section before you install or update vRealize Operations Manager.
Sizing and Scaling
The CPU, memory, and disk requirements that meet the needs of a particular environment depend on the number and type of objects in your environment and data collected, which includes the number and type of adapters installed, the use of HA (High Availability), the duration of data retention, and the quantity of specific data points of interest. VMware updates Knowledge Base article 2093783 with the most current information about sizing and scaling. The Knowledge Base article includes overall maximums and spreadsheet calculations that provide a recommendation based on the number of objects and metrics you expect to monitor.
You can deploy vRealize Operations Manager 6.4 with the following installation formats:
- VMware virtual appliance
- RHEL installation package
- Microsoft Windows installation package Note: vRealize Operations Manager 6.4 is the final version of the product to support Microsoft Windows installations.
If you are deploying a vRealize Operations Manager virtual appliance, use a VMware vSphere Client to connect to a VMware vCenter Server, and deploy the virtual appliance through the vCenter Server instance. The vRealize Operations Manager virtual appliance must be deployed on hosts that ESX/ESXi 5.1 U3 or later managed by VMware vCenter Server 5.1 Update 3 or later.
Note: Although the RHEL-based installation option is fully supported in vRealize Operations Manager 6.4, we will be deprecating this support. The future availability of the RHEL options is not guaranteed.
VMware recommends the use of our Virtual Appliance option rather than using a Linux-based option. The announcement of the End-of-Availability (EOA) for Linux-based options will be provided in a future release.
Linux and Windows Requirements
You can install and upgrade the standalone version of vRealize Operations Manager on the following operating systems:
- Red Hat Enterprise Linux (RHEL) 6.5, 6.6, and 6.7
- Windows Server 2012 R2
- Windows Server 2008 R2 Service Pack 1 (SP1)
Windows Server 2008 R2 Service Pack 1 (SP1) configurations also require the updates found in the following Microsoft Knowledge Base articles:
Applying License Keys
In addition to the existing per operating system instance (OSI) licensing, vRealize Operations Standard edition is also available on a per processor (CPU) basis. Licensing for other editions of vRealize Operations is unchanged and continues to be available on a per OSI metric.
Customers can install multiple Advanced and Enterprise editions in the same vRealize Operations Manager deployment. Customers can deploy single choice and suite licenses together. License counting for individual license keys is handled through licensing groups. You can mix editions or licensing models in one of the following ways:
- Deploy vSphere with Operations Management (any edition) and vRealize Operations Standard together in one deployment.
- Deploy vCloud Suite/vRealize Suite Standard, Advanced or Enterprise and vRealize Operations Advanced or Enterprise edition together in one deployment.
Note: You cannot deploy a vRealize Operations Standard edition license with either an Advanced or Enterprise license in the same deployment.
A license key assigned to a vRealize Operations Manager instance only counts towards virtual machines that are powered on. In order to be compliant, you need enough licenses to cover average powered-on VMs over trailing 365 days.
To ensure compliance of your vSphere 6.0 and 5.5 objects, vRealize Operations Manager 6.4 includes compliance alerts for VMware vSphere Hardening Guide versions 6.0 and 5.5. These hardening guide alerts are based on object type.
In addition to the VMware solutions (vSphere, Endpoint Operations Management, and vRealize Log Insight), see the compatibility table in the Solution Exchange marketplace for many more solutions. These solutions works with Virtual Appliance single or multiple node, RHEL single or multiple node, and Windows single or multiple node configurations.
VMware Product Compatibility
This vRealize Operations Manager release has been certified as compatible with the following VMware products:
- VMware vSphere 5.5 and later
- vRealize Hyperic 5.8.5 and later
- vRealize Infrastructure Navigator 5.8.5 and later
This vRealize Operations Manager release supports all current Web browsers, although only the following browsers have been tested with this release:
- Google Chrome: Version 53 and 54
- Mozilla Firefox: Version 48 and 49
- Microsoft Internet Explorer: Version 11
Installing and Updating vRealize Operations Manager
Note: Customers upgrading older versions of vRealize Operations Manager (6.0.x or 6.1) must use a two-step upgrade: They must first upgrade to vRealize Operations Manager 6.3 and then upgrade to vRealize Operations Manager 6.4. Customers with recent versions of Realize Operations Manager (6.3, 6.2, or 6.2.1) can upgrade directly to vRealize Operations Manager 6.4.
You can also download the following guides before installing or updating vRealize Operations Manager:
- vRealize Operations Manager vApp Deployment and Configuration Guide provides guidance for a vRealize Operations Manager virtual appliance.
- vRealize Operations Manager Installation and Configuration Guide for Linux and Windows provides guidance for vRealize Operations Manager on Red Hat Enterprise Linux or Windows.
Note: Under New Known Issues, see the Installation and Update section for issues related to installing vRealize Operations Manager or updating the software. There are also some older known issues regarding installation or upgrading that may be relavant. Please look in the lower toggled set of notes for these issues.
Resolved Issues From Previous Releases
This section contains the titles of previous known issues that are resolved in this release:
- Inefficient LDAP queries from vRealize Operations Manager cause performance issues on a domain controller.
- vSphere client not displaying badges.
- IPv6 unavailable for DHCP
- Maintenance schedule missing after upgrading to version 6.3
- Launch in Context cannot be removed after deleting vRealize Log Insight server
- Pre-installation checker fails while reinstalling vRealize Operations on the RHEL machine
- Cannot remove a previous master node
- Cannot retrieve Alerts related to removed objects
- Endpoint Operations Management solution plug-ins appear following an unsuccessful solution installation
- UI does not refresh upon login
New Known Issues in This Release
The new known issues that have been identified since the previous release of the product are grouped as follows:
New Installation and Upgrade Issues
- Upgrades from 6.3 to 6.4 may fail.
A PAK upgrade from 6.3 may fail due to a node in a state of waiting for analytics and not responding.
Workaround: Before upgrading from 6.3 to 6.4, customers must take clusters offline.
- Installation fails when $HOME directory has insufficient free space.
Installation fails when the $HOME directory has insufficient free space to complete the installation. You need at least 3GB in
/home/adminor set the IATEMPDIR environment variable to a directory on a disk partition with enough free disk space.
Workaround: Run the following commands on each node and also append them to the /etc/bash.bashrc file:
mkdir -p $tmpdir
- Upgrade endpoint operations agent after upgrading the product.
Customers must upgrade the vRealize Operations product to 6.4 before upgrading the endpoint operations agent to 6.4.
Workaround: Customers should consult KB article 2138748 and KB article 2137709 before starting to upgrade.
- Default policy setting may disable some containers.
In vROPS 6.4, the default policy setting does not enable the memory "Demand" and "Allocation" containers at installation.
Workaround: To enable these containers, do one of these two steps:
- Disable the memory "Consuming" container in the default policy and enable memory "Demand" and "Allocation" containers.
- Go to Administration -> Policies and select the "Default Policy" policy as the system default.
- Inaccurate numbers of member objects in custom groups after upgrading.
The preview count of the members of a custom group is inconsistent with the actual number of members if the group has any Endpoint Operations objects when you upgrade to vRealize Operations Manager 6.4. This will only occur during upgrades, not for new installations.
Workaround: Customers need to use a VMware script to delete the Postgres self relationship for the Endpoint Operations objects. Contact Technical Support to obtain the script.
- Object Summary blank after upgrade.
Object summaries may be blank for all objects if customers do not clean the browser cache after an upgrade.
Workaround: Clean the cache after upgrading.
- The "Reset default content" checkbox in the "Select Software Update" section does not work correctly.
Customers who are upgrading to 6.4 may find that the "Reset default content" checkbox in the "Select Software Update" section, which is a method for overriding the default content that is normally installed during an upgrade, does not work correctly.
Workaround: After the upgrade completes, go to Administration > Solutions, select the vCenter Server solution, and click Modify. Pay attention to upper right corner.
New General Issues
- Log Insight agent hostname reset to default localhost after upgrading
The Log Insight agent hostname configuration will be reset to the default localhost after upgrading.
Workaround: Edit the liagent.ini file to set the hostname to an appropriate value for your environment.
- Some metrics will be duplicated with a label of “old."
in version 6.4 of vRealize Operations Manager, some metrics will be duplicated and will labeled with the suffix “_old.” .
- Slight mismatch in compliance scores on the Compliance tab.
When the vCM management pack is configured, there may be a slight mismatch in compliance scores. For example, the Analysis > Compliance tab may show 46 and the compliance badge may show 47 when the actual score is 46.15.
- Compliance badge score incorrect.
Compliance badge score is 100 for all objects even when compliance alerts are not enabled. By default, the scores should be "?" for all objects in the system.
- Cannot create custom group using adapters that make vRealize Operations Manager REST API calls.
Certain adapters do not invoke the vRealize Operations Manager APIs related to custom groups correctly. As a result, the process to create a custom group in vRealize Operations Manager with objects discovered by these adapters fails with an exception error.
Workaround: The error indicates a problem with the Java language bindings of the vRealize Operations Manager REST API. This currently applies to several vRealize Operations Manager management packs. To work around the problem, see KB 2147670.
- A CSV report file does not include data from a "Dashboard" data type.
When you generate a report that contains a dashboard, the downloaded CSV file includes data of the "Views" data type only.
- Component health status on Self-Health dashboard may not update in a timely manner.
When state of any component of vRealize Operations Manager changes then reflect in health status of respective component may not be shown on Self-Health dashboard for a long time.
Known Issues From Previous Releases
For a list of previously known vRealize Operations Manager issues, click here.
- Installation and Update Issues
- General Issues
- Cluster Management Issues
- User Interface Issues
- Endpoint Operations Management Agent Issues
Installation and Upgrade Issues
- vRealize Operations Manager calculates workloads for data center and custom data center objects differently from cluster objects, basing on provisioned capacity instead of entitlement
Entitlement is the HA capacity of a cluster. vRealize Operations Manager suggests workload placements across clusters based on the entitlement provisioned to each cluster, which depends on whether HA is enabled or disabled on the cluster.
- When HA is disabled, the entitlement is equal to the hardware provisioned capacity of the cluster. The workload is equal to (Demand + Overhead) ÷ (Provisioned Capacity).
- When HA is enabled, the entitlement is equal to the sum of the cluster overhead and the currently computed HA capacity, minus the HA reservation. The workload is equal to (Demand + Overhead) ÷ (Entitlement).
- Some object alerts not enabled after upgrading to the current release of vRealize Operations Manager.
After users upgrade to version 6.3 or 6.4 from an earlier version that had vSphere Hardening Guide alerts enabled, users will not see alerts related to distributed port groups, distributed virtual switches, or vCenter objects as these alerts are not enabled by default on upgrade.
Workaround: Choose one of the following:
- Enable hardening guide alerts from the adapter configuration.
- Go to the policy library, edit the active policy to enable alerts for these three objects.
- vRealize Operations Manager alerts trigger when the Management Pack for vCenter Configuration Manager (VCM) is integrated
When the Management Pack for VCM is installed and the vSphere Hardening Guide alerts are enabled for 5.5 or 6.0 objects, some of the native compliance alerts inadvertently trigger in vRealize Operations Manager. The compliance score for a selected object displays the violated native compliance rules in the Compliance tab, which results in an incorrect compliance badge score.
Workaround: Disable the native compliance alerts with these steps:
- To see the native compliance alerts that are enabled in the policy, click Content > Alerts.
- Click Administration > Policies > Policy Library.
- Edit the policy applied to the objects and disable the native compliance alerts.
- Watchdog might restart Tomcat
If there are multiple agents registered to a master node directly, then Watchdog might restart Tomcat whenever there is a need, where UI runs on it. There is, however, no such implication on the product or the cluster.
Workaround: Register agents to remote collector or a load balancer instead of to a Master node.
- Non-ASCII Windows user account cannot install
Installation of vRealize Operations Manager on Windows fails if the logged in user has an account name that contains non-ASCII characters.
Workaround: Before installing on Windows, log in with an account name that only contains ASCII characters.
- Removal of directories might affect upgrade
In some cases, users have manually removed vRealize Operations Manager directories, such as directories where logs are stored. For example:
The upgrade process requires vRealize Operations Manager directories to exist. If the directories are missing, the upgrade might fail by preventing vRealize Operations Manager services from starting, or in other unpredictable ways.
Workaround: Do not manually remove vRealize Operations Manager directories.
- It might be unclear that an update has failed
If a customer reboots a node or kills some active processes during a software update, the update process continues to show that the update is still underway even though it has failed.
- Residual files prevent reinstallation on RHEL and Windows
After you uninstall vRealize Operations Manager from Red Hat Enterprise Linux (RHEL) or Windows, files such as logs might not be completely removed. If residual files exist, reinstallation might fail.
Workaround: Before reinstalling, manually remove residual files and directories from the vRealize Operations Manager install and data directories. The following are the default directories:
- RHEL uninstallation does not fully remove user accounts
After you uninstall vRealize Operations Manager from Red Hat Enterprise Linux (RHEL), vRealize Operations Manager admin and postgres users still appear in the /etc/passwd file.
Workaround: None. You may manually remove the users from the file if desired.
- RHEL uninstallation does not uninstall Postgres
After you uninstall Red Hat Enterprise Linux (RHEL), Postgres is not uninstalled. The following are left behind:
Workaround: Users must manually remove these files.
- The vCenter Solution for Endpoint Operations installation fails when the vRealize Operations Manager inventory has objects discovered by previous versions of the vCenter Solution for Endpoint Operations
If you have any object collected from earlier version of the vCenter Solution for Endpoint Operations, then ensure that the objects are removed before you install or upgrade to the latest version of the vCenter Solution for Endpoint Operations.
- Stop the collection of all objects or resources that are discovered by the earlier vCenter Solution for Endpoint Operations.
- Delete all the stopped objects.
- Install the latest vCenter Solution for Endpoint Operations.
Note: By removing the objects, you will lose historical data from the vCenter Solution for Endpoint Operations and vCenter application objects history.
- Generated reports may be blank
If you log in to vRealize Operations Manager with vCenter Server credentials and generate a report, the generated report is always blank.
- Endpoint Operations Management does not support IPv6
Endpoint Operations Management does not support IPv6
- VVOL datastores are masked
If a VM, host, cluster, datacenter, or vCenter Server object with VVOL datastores is selected, an incorrect datastore count appears on the Environment Objects Map or the Troubleshooting All Metrics page. And if a VM is deployed on a VVOL datastore, raw or derived metrics may be missing.
Workaround: None. The VVOL datastores are masked because vRealize Operations Manager does not yet implement the VVOL datastore metrics. However, the VVOL datastores are included in the count.
- Rebalance Container action continues to recommend the same rebalance plan after failure
When a Rebalance Container action fails, and vCenter Server indicates that the operation is not allowed in the current state, the rebalance container action recommends the same rebalance plan.
Workaround: Correct the problem in your environment, then attempt to run the Rebalance Container action again.
- Maintenance schedules no longer assigned after upgrade
Maintenance schedules are disassociated from a policy after an upgrade from version 6.0 to later versions.
Workaround: Update the policy manually after an upgrade. For more information about managing policies, see Managing and Administering Policies for vRealize Operations Manager.
- Host Maintenance mode behavior has changed in vRealize Operations Manager 6.x.
Now when a vCenter host is in Maintenance mode, the system does not disable the data collection icons from the host. Instead, in vRealize Operations Manager 6.x, maintenance is governed by policies and associated with custom groups.
Workaround: You can create a custom group for hosts that are in maintenance mode and apply whatever policy you want for them, including stop collection of all metrics.
- Deploying virtual machines to a failover host is not allowed
After you configure a failover host in a cluster and select the failover host in the navigation tree, then go to Analysis > Capacity Remaining, create a custom profile for virtual machines, and enable the custom profile for all policies, you can view the number of virtual machines that will fit on the host.
Although the custom profile indicates that a number of virtual machines will fit on the host, you cannot power-on virtual machines deployed in failover hosts.
- Symptom set details cannot be retrieved for some alerts
When an alert triggers based on a set of population-based symptom sets, and one or more symptom set could not be evaluated due to lack of enough resources to evaluate the population condition, an error occurs. The error message indicates that the complete symptom set details for that alert cannot be retrieved for that particular symptom set. This error only occurs on population-based alerts and only if enough resources were not available to evaluate the condition.
- Alert Details summary links below the top link do not link to objects
When you view the alert details for an alert that has received symptoms from many objects, such as an alert on a host that has symptoms from virtual machine child objects, and then you expand the drop-down arrow for the alert cause, you can see the list of symptoms. Because the alert is generated with symptoms received from many objects, you see a list of links to those objects with relevant symptom descriptions. The links to the objects located below the first object do not work. This problem is relevant for all browsers, except for Internet Explorer.
Workaround: Manually find the object you need. For example, use the global search text box in the top toolbar to search for the object.
- Virtual Machines will not Power On, even though vRealize Operations Manager reports capacity remaining
In cases where one virtual machine is larger than the others in the cluster, the size discrepancy from this single virtual machine can result in vRealize Operations Manager reporting that there is capacity and time remaining, yet Admission control in vSphere will not allow the virtual machines to be powered on.
Workaround: Determine the slot size and then increase the buffer threshold using policies. See KB 2033248 for more information about this issue and how to manually calculate the slot size.
- Workload placement might fail if the ESXi host version is missing
If vCenter Server fails to provide the ESXi host version number to the vCenter Server properties collector in the vCenter Adapter, a compatibility problem occurs, which can cause the Rebalance Container action and the workload placement to fail. This failure might persist until the vCenter Adapter establishes a new connection to the vCenter Server instance.
Workaround: Restart the vCenter adapter instance. In the vSphere Web Client, view the summary and version information for the ESXi host machines.
- Some symptoms for metrics are not canceled for objects in an unknown or no-data-received state
When vRealize Operations Manager stops collecting new values for a metric, vRealize Operations Manager cancels symptoms when the last collected value of a metric becomes stale. In some cases, a symptom is not canceled. For example:
- When an adapter is stopped, vRealize Operations Manager stops processing its objects, including the cancelation of stale symptoms.
- When a vCenter Server instance is deleted and the adapter instance is not deleted, vRealize Operations Manager stops processing its objects, including the cancelation of stale symptoms.
- The configuration for the content pack for Log Insight is overwritten
When you change the configuration details in the loginsight-agent file or from the user interface of vRealize Operations Manager, the configuration for the content pack for Log Insight is modified or overwritten. The ports from 9000 to 514 might be reset and the protocol changes to syslog instead of cfapi.
Workaround: Disable the remote logging option from the vRealize Operations Manager user interface before you modify the loginsight-agent file. Do not save the changes again from the user interface when you use the content pack for Log Insight.
- The Text Display widget cannot display some Web sites in HTML mode
If a Text Display widget is configured with the View Mode set to HTML, then the URL should use HTTPS protocol. The functionality of the Text Display widget with Web sites that use HTTP depends on the individual settings of the Web sites.
- Projects and workload placement do not acknowledge disk I/O and network I/O
Projects do not impact disk I/O and network I/O metrics. The placement plan might include placing virtual machines on hosts with high disk I/O or network I/O workload, or datastores with high disk I/O workload.
- When you do a rebalance on a container with hosts of different capacities, the balance might not be achieved in terms of usage and demand percent
When you do a rebalance on a container, host candidates with larger absolute available capacity is considered first to place virtual machines. As a result, when hosts are of different capacities, balance may not be achieved in terms of usage and demand percent.
- The Network Share plug-in cannot work with path that contains non-ASCII characters
If the path in your Network Share plug-in contains non-ASCII characters the connection might fail.
Cluster Management Issues
- Cannot add nodes that are on earlier version than cluster
Once you have updated a cluster to vRealize Operations Manager 6.2.1 or later, you cannot add a node that is on an earlier release version.
User Interface Issues
- The filter menus in Policies workspace step Collect Metrics and Properties do not work correctly
When you deselect an option from the State, KPI, or DT drop-down menus, the filter hides all attributes no matter of their settings.
Workaround: Save and reopen the policy.
- A PDF report file might not include all data from a widget that has a scrollbar
When you generate a report that contains a custom dashboard with a widget that has a scrollbar, the downloaded PDF file might not include all data of the widget.
Workaround: Increase the height of the widget in the dashboard to fit all the data.
Endpoint Operations Management Agent Issues
- Blank password not accepted for Postgres user
The embedded Postgres database for a vCenter Server appliance is shipped without a default password for the postgres user. If you do not set the password in the vcdb.properties file, you are unable to log in as the Postgres user because the Endpoint Operations Management agent will not allow you to log in with a blank password. You might experience this problem if you are trying to monitor the Postgres database after importing the Endpoint Operations Management Adapter.
Workaround: Use the Postgres DB credentials stored in
/etc/vmware-vpx/vcdb.propertiesfor the Virtual Appliance and
C:\ProgramData\VMware\vCenterServer\cfg\vmware-vpx\vcdb.propertiesfor Windows. It is not necessary to enter the default (blank) Postgres password.
- vRealize Operations Manager Monitoring a Large Scale Environment is not Performing as Expected
If you have a large scale vRealize Operations Manager monitored environment, you may experience one or more of the following issues:
- The user interface is responding slowly
- Metrics are slow to be displayed
- Roll back of Windows Endpoint Operations Management agent service upgrade does not roll back JRE
If you upgrade a Windows Endpoint Operations Management agent service to version 6.2 and the upgrade fails, the upgrade rolls back to version 6.1. However, the JRE that was upgraded as part of the upgrade process does not roll back to its earlier version.
Workaround: Restart the agent service. The earlier version of the JRE is invoked.