vcenter Operations Management Pack for Storage Devices Release Notes vcenter Operations Management Pack for Storage Devices Build number1450289 Last document update: 16 Dec 2013. Contents: Compatibility and Interoperability Known s Management Pack for Storage Devices provides you visibility from virtual machine to storage LUN, end-to-end of the storage area network I/O. This allows you to monitor and troubleshoot capacity and performance issues on different components of your storage area network. Compatibility and Interoperability Management Pack for Storage Devices is compatible with vcenter Operations Manager 5.8. Supported SMI-S versions : 1.1.0 or 1.5.0 Supported switch: Brocade with BNA (Brocade Network Advisor) 12.0.1 or greater Supported firmware : Brocade 4100 : v6.4.3 or greater; Brocade 6510 : v7.0.1a or greater Management Pack For Storage Devices does not automatically detect vcenter servers configured in linked mode. If your implementation requires vcenter to be configured in linked mode, then you must create an adapter instance of Management Pack for Storage Devices for each vcenter in your system. Supported ESX versions : ESX 5.1 U2 and above Known s The resolv.conf is used by vcenter Operations Manager virtual machine operating systems to connect to remote machines over the network. If this file is not properly configured, then vcenter Operations Manager might not be able to reach destination machines (for example, VC, ESX hosts or FC switches). Specify the search domain value on the Analytic and UI virtual machine to reach vcenter, hosts, and switches using their DNS names instead of IP addresses. Specify the search domain in the file /etc/resolve.conf on the Analytics and UI virtual machines, then restart the collector. See http://en.wikipedia.org/wiki/resolv.conf. To specify the search domain: 1. ssh as admin to the Analytic and UI virtual machines. 2. Edit the file /etc/resolv.conf 3. Specify the search domain (and nameserver) 4. After completing this change, restart the collection service: sudo -s service vcops restart collector Resources deleted manually from the vcenter Operations Manager UI are not rediscovered automatically.
If you manually delete resources such as HBAs, switchports, target ports, and zones from the vcenter Operations Manager UI, they might not be rediscovered automatically. Restart the vcenter Operations Manager collector to rediscover resources that are connected. vcenter Operations Manager collection service fails with out of memory errors s The permanent memory (PermGen) allocated to load collector JAR files is too small for the number of adapters in the deployment. To resolve this issue, increase the amount of memory allocated to PermGen: 1. ssh as admin to the Analytics virtual machine. 2. Edit the file /usr/lib/vmware-vcops/user/conf/collector/wrapper.conf. 3. Increase the MaxPermSize parameter. For Management Pack for Storage Devices, change the line wrapper.java.additional.3=-xx:maxpermsize=128m to: wrapper.java.additional.3=-xx:maxpermsize=256m After completing this change, restart the collection service: sudo -s service vcops restart collector Target ports for an array are not showing on troubleshooting dashboards. VASA Provider/credentials are not available in vcenter. Add credentials in vcenter and restart the collector. Certain supported hosts and switches are not visible in the dashboards. Check the error log for incorrect vcenter/fabric credentials. For example: ERROR [Collector worker thread 1] (115) com.integrien.adapter3.vsom.datasource.vcdatasource.connect - SMS connection could not be initiated com.vmware.vim25.invalidloginfaultmsg: Cannot complete login due to an incorrect user name or password.
Provide the correct credentials and save them. Please wait for a few collection cycles to see the newly discovered resources on the dashboards. The Heatmap dashboard displays uncolored boxes. The corresponding resources might no longer be connected to the vcenter Operations visible infrastructure. Resources from unsupported hardware also show up as uncolored boxes on dashboards. Delete the removed/unsupported resources from the Environment Overview and restart the collector. Dashboards are not visible after installation on Windows/Linux. Post install scripts need to be run on standalone installations. See "Install Management Pack for Storage Devices Dashboards in a Standalone Environment" in the Installation and Configuration guide. Can I register the switch IP address in the credentials field? It is scalable and reliable to add the IP address of the host (on which BNA (Brocade Network Advisor) is installed) as credentials. VMware recommends that you use BNA to connect the switches to Management Pack for Storage Devices. Note that firewall ports might have to be opened (or firewalls disabled) on the machine that has BNA installed to allow BNA to communicate successfully with Management Pack for Storage Devices. The following exception repeatedly appears in logs: ERROR [pool-6-thread-7] (4757) com.integrien.adapter3.vsom.datasource.esxdatasource.isconnected - Connection Object not valid WBEMException: CIM_ERR_ACCESS_DENIED (HTTP 401 - Unauthorized) The exception is logged when the ticket connecting vcenter to ESX expires. The exceptions are harmless and can be ignored.
A virtual machine on a datastore using two or more LUNs appears to be connected to fewer LUNs in the Troubleshooting dashboard. The dashboard is customized to show only the LUNs in use by the virtual machine. This information is updated every collection cycle. Alerts such as an FC link going down on switchports are not shown in real time on the dashboards. Alerts are updated every collection cycle. Alerts do not appear on the Alerts dashboard after several collection cycles pass. This can sometimes happen if similar alerts are posted to the resource previously. The older alerts need to be cancelled before new alerts can be posted on the same resource. Cannot locate logs related to Management Pack for Storage Devices. The logs for the Management Pack can be accessed under /usr/lib/vmwarevcops/user/log/adapters/universalstorageadapter3/universalstorageadapter3*.log in the Analytics VM. The collector logs also contain log statements for MPSD, labeled with com.integrien.adapter3.vsom. Logging level for for INFO and DEBUG is not showing after updating log4j.properties. Restart the collector after updating log4j.properties. A CIM provider on ESX is unable to provide information to Management Pack for Storage Devices. When you run the command # enum_instances IODM_SCSIadapter you receive the error: error: enuminstances ThreadPool --- Failed to enqueue request. Too many queued requests already: iodm_group, active 5, queued 11 The CIM provider on ESX is unable to keep up with the number of requests coming from MPSD. You must increase the time interval between collection cycles for Management Pack for Storage Devices in vcenter Operations Manager.
To increase the time interval between collection cycles: 1. Log in to the vcenter Operations Manager UI as admin 2. Select Environment > Environment Overview 3. Select Resource Kinds > Storage Devices Instance > All attributes 4. Select the Management Pack for Storage Devices instance in the right pane and click Edit 5. Increase the value in the Collection Interval field. 6. Click OK. Note: Additional memory provided to the ESX CIM provider reduces free memory that might be available to virtual machines running on ESX. To increase the number of requests ESX CIM provider can service: 1. ssh into ESX server. 2. Edit /etc/sfch/sfcb.cfg. 3. Change the values as needed for the following parameters: threadpoolsize (default 5, minimum 3, maximum 25) requestqueuesize (default 10, minimum 0, maximum 30) intsocktimeout (in seconds, default 600). 4. Restart sfcbd with the command /etc/init.d/sfcbd-watchdog restart. Note: It might take several minutes for sfcbd to restart. Clicking on a virtual machine resource in the Troubleshooting dashboard shows an incorrect set of target ports in use (in addition to the correct target ports). A host can have multiple adapters with each adapter connected to a single switch. This switch might be connected to target ports on different arrays. Since the health of the host (and the virtual machine on the host) can be affected by the health of the target ports to which it has access, clicking on a virtual machine (or host) resource in the Troubleshooting dashboard shows all target ports that the host can reach.