Just a few months after the 6.1 release, the product team has released vRealize Operations 6.2. You can find many articles about it by bloggers, such as one this and this. Someone going with the YouTube ID of VirtualXpress IN has even done a video on the upgrade.
Check the known issue here. Also, check the relevant KB, such as this one. For my case, the disk space looks relevant, so I ensure I have enough space first. There are a set of issue on cluster management, none relevant for me as my customers find vSphere HA as sufficient. For an environment with <5000 VM, I think a single node is sufficient. Speaking of high availability, I’d rather have 1 vRealize Operations VM protected by SRM then 2-node cluster without SRM.
Take note of the “Redirecting syslog files to the Log Insight feature also redirects garbage collection files” issue. While I have sufficient space in my Log Insight instance, I do not want it to be cluttered with information that I do not need.
There is an issue with Report Template tab toolbar in using IE 11. In general, I avoid IE for vRealize Operations as I’ve seen the performance delta.
The Release Notes provides tips that if you have a remote environment in which the latency is > 20 milliseconds, you should install a remote collector. I prefer to deploy an appliance instead of Windows as remote collector so I can stick to just 1 common platform for vRealize Operations.
Here are other known issue from the Release Notes that I’m taking note, just in case I see them in my engagements.
- Some vSphere 6 diskspace metrics do not appear
Only the following diskspace metrics are published to vRealize Operations from vSphere 6.
Virtual machine: diskspace|used, diskspace|diskused, diskspace|snapshot
Host: diskspace|used, diskspace|diskused, diskspace|snapshot
Cluster: diskspace|used, diskspace|diskused, diskspace|snapshot
Datastore: diskspace|used, diskspace|diskused, diskspace|snapshot, diskspace|swap, diskspace|capacity, diskspace|disktotal, diskspace|provisioned
- Metric for Outstanding I/O requests might be inaccurate on some object types
The datastore|demand_oio metric that appears as Datastore|Outstanding I/O Requests in the UI, might be inaccurate for the vCenter Server, Datacenter, Cluster, and Datastore object types. Due to the method used to calculate a weighted average of transactions and the latency of transactions across hosts, the value of this metric might be artificially high.
- Deselecting Read permissions for Resources and Dashboards has no effect
When you create a user role with no Read permission for Resources and Dashboards, the user can still see the dashboards and all resources on the home page. Object Relationships page is always visible to the user with or without any permission.
- A PDF report file might not include all data from a widget that has a scrollbar
BTW, I’m upgrading from 6.1 Patch. I no longer have earlier edition running around 🙂
The screenshot below is to show you that the upgrade process picked up when I forced stop the upgrade process as I thought it hang. Glad to know it was able to pick up! I clicked Finish installation button.
It took me to this screen, which is by now a familiar screen if you have done multiple update. The keyword here is OS. You need to update the SuSE Linux first. Tomas Baublys wrote a good guide here.
If you leave the browser screen alone, it will timeout after several minutes. Don’t worry, you can log back in and see that upgrade is on-going.
In my case, it was just a 5-10 minutes, and it’s done. This is what it looks like.
Once the base SuSE Linux is updated, time to update the vR Ops application. I shared the process to update from version 6.0 to version 6.0.1 here, and from 6.0 to 6.1 here. The process to update to 6.2 is the same, so kindly review the above. No need to fill the Internet with duplicate info 🙂
11 February 2016 update: There is a patch release called 6.2.0a. See the release notes here.