VMware

vCenter Hyperic 5.8.2 Release Notes

vCenter Hyperic 5.8.2 | 24 June 2014

Last Document Update: 21 July 2014

The VMware® vCenter™ Hyperic® 5.8.2 release notes cover the following topics:

What's New in vCenter Hyperic 5.8.2

In addition to general issue fixing, the following changes have been made in vCenter Hyperic 5.8.2.

  • vCenter Hyperic version supports running the vCenter Hyperic agent on Oracle Solaris x86.
  • To enhance security, the embedded PostgreSQL database has been removed from the Windows installation.
    Customers with an existing embedded PostgreSQL database who want to upgrade to the new version must export their existing data and import it to a standalone PostgreSQL database.
  • The vCenter Hyperic installers are now using JRE 1.7_51.

Documentation for the vCenter Hyperic 5.8.2 Release

The documentation for vCenter Hyperic 5.8 is applicable to the vCenter Hyperic 5.8.2 release.

Issues Addressed in vCenter Hyperic 5.8.2

  • The issue in which network device services were not discoverable by the vCenter Hyperic agent has now been addressed.
  • The issue in which the availability of a platform in the metric viewer showed 200% has been addressed.
  • The issue in which the triggering of a Config changed alert caused corruption of the resource-specific Alert tab has been addressed.

Known Issues and Limitations in vCenter Hyperic 5.8.2

The following known issues and limitations affect the vCenter Hyperic 5.8.2 release.

  • New! This issue is relevant when you are using a new installation of vCenter Hyperic 5.8.2, not an upgrade from an earlier version. If you are using vCenter Operations Manager with the vCenter Hyperic Management Pack, a restart of either the vCenter Operations Manager or the vCenter Hyperic Management Pack causes duplicate vCenter Hyperic objects to be created in vCenter Operations Manager. This situation occurs because each time that a get GUID query is called, a different GUID is returned, which leads to creation of the duplicate.
    Workaround
    1. In vCenter Operations Manager, select Environment > Environment Overview.
    2. Open the Adapter instances > MP for Hyperic Instance directory and select the instance of your Hyperic 5.8.2 installation.
      A table appears, displaying the vCenter Hyperic resources and their data.
    3. Sort the table according to the last Creation Time (Descending).
    4. Copy the value in the Identifier 1 column of the first Hyperic resource row. This is the value of HQ-GUID ID that you will paste later in the procedure.
    5. Run psql -U [HQ DB username] -d [Database Name] -W to log in to the PostgreSQL server as the HQ DB user.
      • You can identify the HQ DB user name from the hq-server.conf file. In the server.database-user=hqadmin property, hqadmin is the username.
      • You can identify the database name from the hq-server.conf file. In the server.database-url=jdbc:postgresql://127.0.0.1:5432/HQ?protocolVersion=2 property, HQ is the database name.
      • Use -W to ensure that the user is prompted for the database password HQ DB Username.
    6. Back up the vCenter Hyperic database.
    7. Update the HQ-GUID in the following update query by pasting the HQ-GUID that you copied earlier in the procedure and then update the database by running UPDATE EAM_CONFIG_PROPS SET propvalue = "<HQ GUID ID>" WHERE propkey = "HQ-GUID";
    8. Run commit; to commit the database.
    9. Restart the vCenter Hyperic server.
  • New indicators have been added to the System plug-in. When you perform a version upgrade, the new indicators do not display data. This issue affects the Total Memory, Used Memory, and Swap Total metrics that are now converted to indicators, and the two new metrics Total Disk Capacity and Total Disk Usage.
    Workaround:
    Either
    1. Define a policy that includes all the default metric configurations.
    2. Create a dynamic group that includes all the platforms.
    3. Assign the policy that you defined to the dynamic group.
    4. The metrics update and are displayed as expected.
    Or
      For each platform resource, on the Metrics tab go to Metric Data > Show all Metrics and configure an interval for the missing and empty indicators.
      Following the interval configuration, the indicators display the metrics.
  • When using an Oracle database, in the event that the listener is down, the Start action only works if the vCenter Hyperic agent is running using the same Windows user credentials as those that are being used for Oracle.
  • The vCenter Hyperic agent monitors an Oracle database by maintaining an open connection to the database. If the listener goes down, but the Oracle database instance is still up, vCenter Hyperic continues to detect the instance as available, because the existing connection remains open. However, new connections cannot be made because the listener is down, meaning that the database is not actually available.
    Workaround Restart the hq agent. The Oracle database availability becomes red, and the control actions work.
  • In Administration > Edit HQ Server Settings, if you incorrectly configure credentials under the vCenter Settings, you are not alerted that you have provided invalid credentials but vCenter mapping does not function.
  • Nltest only runs on Microsoft Exchange 2007 installed on Windows Server. You must run vCenter Hyperic as a domain user of Microsoft Exchange, not the default local system user that is specified during vCenter Hyperic installation.
  • After upgrading to vCenter Hyperic 5.8, mysql-plugin.jar continues to appear in the Plug-in Manager, even though it no longer exists.
  • Occasionally, when the vCenter Hyperic agent is down, the Resource Health status appears as down, when the actual status is unknown.
  • When you rename a resource, you must update the group before the change is visible to the group.
  • After you have installed a vCenter Hyperic agent on a Windows virtual machine, the agent starts automatically.
  • When you upgrade a vCenter Hyperic agent, if you select an option that is incompatible with the upgrade, the agent becomes unavailable and the displayed messages are misleading.
    Workaround:
    1. Open the rollback.properties file to roll back version 5.8 to 5.0.
    2. Specify set.HQ_AGENT_BUNDLE=agent-5.0.0
    3. Specify set.HQ_AGENT_ROLLBACK_BUNDLE=agent-5.0.0
    4. Restart the agent.
    Note that, although the vCenter Hyperic agent version is version 5.0, under Administration > HQ Health > Agents, the version number still displays 5.8.
  • The Add/Update Plugin(s) button does not apply updates to plug-ins, it only adds new plug-ins.
    Workaround:To update a plug-in, go to Metric Data > Show All Metrics and set the interval for any metric. The agent will collect the new metrics.
  • If you change the pattern property in an HTTP resource, the change does not take effect.
    Workaround: Delete the HTTP resource, and recreate it with the new pattern.
  • The default vCenter Hyperic installation directory on Windows is C:\Program Files. The vCenter Hyperic installer does not recognize spaces in directory names.
    Workaround: Change the default installation directory to one with no spaces in its name.
  • A user who is the owner of a policy cannot be deleted.
    Workaround: Delete the policies that are owned by the user, before deleting the user.
  • An Availability metric for platforms that is set for one minute must be treated as a minimum of two minutes when you choose the recurrency option for an alert definition.
  • The Exchange 2007 standalone plug-in must be installed in Windows server 2003.
  • In the case in which multiple vCenter servers contain virtual machines that share the same MAC address, if some of the vCenter servers are configured to vCenter Hyperic server, or have vCenter Hyperic agents running on them, the vCenter Hyperic server might interpret and display those platforms as a single virtual machine.
    Workaround: Configure all the vCenter Servers to the vCenter Hyperic server. A vCenter Server mapping mechanism is generated that removes the MOREF data from the virtual machines with a common MAC address.
  • When you attempt to install the vCenter Hyperic plug-in on an Exchange machine, the server might be unavailable and the message Error retrieving value: \MSExchangeIS Mailbox(_Total)\Messages Sent might be displayed.
    Workaround: Go to http://technet.microsoft.com/en-us/library/aa995801%28v=exchg.80%29.aspx and follow the directions to enable the MSExchangeIS object in the Performance Manager.
  • A virtual machine that does not have VMware Tools installed is not mapped to the vCenter Hyperic server.
  • When you install vCenter Hyperic usings the non-JRE Windows installer, it might take some time before the user interface appears, and invalid exceptions might appear in the log.
  • Deleting a resource from the inventory causes any resources currently in the auto-discovery queue to be removed from the queue and from the Auto-Discovery portlet on the dashboard.
    The resources reappear in the portlet when they are next auto-discovered and added to the queue.
  • In a large vCenter Hyperic deployment, deleting multiple platforms simultaneously from the user interface can result in orphaned resources, browse errors, incorrect results in the user interface, and under some circumstances, a stack trace.
    The following message in server.log indicates that this error condition has occurred:
    05-04-2012 13:10:54,185 EDT ERROR [tomcat-http--19] [org.hyperic.hq.ui.util.UIErrorHandler@49] An exception was thrown for the following request URI
    '/resource/group/monitor/Visibility.do':
    java.lang.NullPointerException
    at org.hyperic.hq.bizapp.server.session.MeasurementBossImpl.findGroupCurrentHealth (MeasurementBossImpl.java:3204)

    If you experience this problem, contact VMware Global Support Services.
  • The Operations Center page intermittently is empty and displays Sorry, an error occurred.
    A stack trace is written to server.log when the error occurs.
  • If a vCenter Hyperic agent is deployed on an operating system with an IP address that is currently associated with an existing resource (platform) in vCenter Hyperic, the agent will not auto-discover the platform, servers, and services associated with this operating system.
    Workaround: Prior to deploying the application, remove vCenter Hyperic resources that are currently associated with the IP address to be reused.
  • When adding alert definitions that include a recovery alert to a plug-in JAR file, the loader fails to parse the recovery alert.
    Workaround: Comment out the recovery alert and the loader will parse correctly.
  • Role names and group names in vCenter Hyperic must be prefixed by org/. For example org/Admin.
  • Do not start or restart the vCenter Hyperic server using root credentials.
  • When SELinux is enabled, the CentOS 5.8/5.7 32-bit vCenter Hyperic Agent running on Java 7 fails to start, and the message failed /work/hyperic-hq-agent-5.8.0/bundles/agent-x86-linux-5.8.0/jre/lib/i386/client/libjvm.so, because /work/hyperic-hq-agent-5.8.0/bundles/agent-x86-linux-5.8.0/jre/lib/i386/client/libjvm.so: cannot restore segment prot after reloc: Permission denied is displayed.
    Workaround: Attempt to run chcon -t texrel_shlib_t /work/hyperic-hq-agent-5.8.0/bundles/agent-x86-linux-5.8.0/jre/lib/i386/client/libjvm.so so that the agent and SELinux run on the same virtual machine.
  • When using Internet Explorer, you cannot customize SNMP trap OIDs for escalation actions and SNMP notifications.
  • Emails are not sent when vCenter Hyperic is deployed from an OVA file.
    Workaround: Update the email configuration properties on the Administration page.
  • In the HQ API, when you call get resource config, the platform.config_track.files value is empty.
  • Sometimes, an incorrect DSN for collecting virtual machine or ESX server data is passed by the server to the vCenter Hyperic agent. In this case, the virtual machine or ESX server appears to be unavailable.
    Workaround: Change the default collection interval for the affected resources, so that the server sends the DSN to the agent again.