Upgrading vC Ops 5.8.3 to vRealize Ops 6.0

Hello,

This is quite an upgrade – in fact it is not an upgrade but a migration.  Meaning you install vRealize Operations 6.0 and then migrate your vC Ops data to vRealize Operations (vR Ops).

This will not do vR Ops in an HA configuration.  In other words it will be the same as vC Ops in terms of HA. However I would love to move to HA if it made sense in my situation.

I am not going to detail this out as it has been detailed elsewhere but I am going to pull it all into here and make sure it works.

This article will get vR Ops working after a migration but will also cover off some other things that are important that may or may not be configured for you – SMTP, Active Directory Integration, and Management Packs for example.

Learning

Here is a number of things that you should check out.  Or skip over them and continue reading!

Release NotesDocumentsIntro 1 and Intro 2.  Here is an article with good architecture info.  You can see a good video of the UI here.  I have noticed that the vR Ops release notes has been updated at least once since it was published.  There is a lot of good info starting around page 70 in this guide.  Here is an interesting sizing document from the KB.  As well, here is a link to a KB article on Management Pack compatibility.

Prerequisites

You need to make sure that your vC Ops is 5.8.1 or later.  In my case for this it is 5.8.3.  It appears that you might have issues with custom dashboards not being migrated if you are at 5.8.2.  Since I am doing an upgrade of 5.8.3 maybe you should upgrade to it first.

  • I like to suggest you make sure your vC Ops is running, and you have a backup – just in case.
  • Don’t be using DHCP for your vC Ops.  If you are change to static – more info here.
  • Create a DNS entry for your vRealize Ops instance.  Have the FQDN and IP address ready – no need for an IP Pool.
  • I suggest you use a FQDN for the name of the VM when deploying the .ova.
  • You will need your certificate handy if you are not going to use the default.
  • You will need your NTP info during the install / config.
  • While vR Ops has a product evaluation period I would suggest you have the vR Ops product key handy.  It is a new one and not vC Ops.

Upgrade Process

There is a number of things to be aware of and accomplish so we will work through it all starting here!

First Install

You need to first install vRealize Ops and you can use the info right here.  Make sure to only do part 1 and NOT do any of the other parts.

Note, install any third party Management packs in vR Ops now.  If you do not, while the migration process will migrate any dashboard, third party data will not be migrated unless you have the MP installed in the destination vR Ops before migration.

Migration

So now you should connect with your vR Ops server and follow this article.  Be aware that the migration agent install may take a few minutes, and add the migration source takes a bit too.  For my small lab it took about 10 – 14 minutes.  I love the info on disk space / CPU and Memory after the migration.  Very handy.  The actual migration time was around 25 minutes.  While you are watching the migration occur, you can click on the import source and see what is going on.

Test

You should make sure you can still see the past – select a host or VM and see what you see.  There should be history of some sort.  Not sure if there is a delay or not.  It appears there is a small one – at least for me.

Additional Configuration

This is a selection of things that I think most users will need to be aware of and complete.

Fix vSphere Web Client Health State Widget and Misc. vCenter communications issues

Check this out.  Don’t skip this!

Log forwarding

It is very nice you can do this so easily.  Find out here.  Be aware that you can find the Logs location referenced in that article by starting at Home, selecting Administration, Support, and then you will see logs.  If you are using Log Insight 2.5 you can very easily – from with LI – download the vRealize Operations Content Pack – which I just checked and it is not there yet.  Only the vC Ops but I am sure soon it will be there.

Important note: After you get log forwarding from vR Op to your syslog server – in my case I use Log Insight, you will not be able to find events from it if you filter using hostname (such as hostname contains vrops_FQDN).  Which is of course how I always do it.  Not sure why but you need to use source to filter.  I wonder if this has to do with something for when it is in a cluster?  Also, sometimes, you will have a hostname of – and that is the vR Ops server. Frustrating. Update – 7/7/15 – this issue doesn’t occur in 6.0.2.

Update: William had done a script to make doing the log forwarding much easier.

Management Packs

Management Packs (MP) are how additional functionality is provided to vR Ops.  Sometimes it might be storage related, or it could be product related.  Generally there will be several you want to install.  Log Insight (LI), VMware Infrastructure Navigator (VIN) and the Storage MP.  You may have third party MP’s as well.

You can find info on installing the three VMware MP in this article.

Active Directory

So much easier to use one set of credentials and most customers I know have selected Active Directory to hold their credentials.  Find out how to connect AD, and define some of your own admin users here and read only here.

Outbound Email Alerting

This is important and you can get help getting it going here.

More Info

  • What migrates from vC Ops 5.x to vR Ops 6.x – here
  • Create a vRealize Ops cluster – here

Things to watch out for

How long do we need to leave the old vC Ops in place?  Can we turn it off right after the migration is done?  You should turn off vC Ops as soon as you are comfortable that vR Ops is collecting successfully and displaying successfully.  Don’t want to run with both as that is double the load on vCenter.

I also had no vSphere Web Client Health Widget info but you should if you followed the steps above.

After several hours I noticed that all of the different solutions – like vSphere – were all still receiving data.  This would lead to incomplete data in the vR Ops UI so hopefully it finishes soon.

As I mention in the Management Pack install article it appears that only VMware Infrastructure Navigator automatically recognized applications – meaning that you define – are going to arrive in vR Ops (update: info and fix here).  Plus, it appears that you get the two VIN dashboards only if you log in as Admin and not if you log in with admin rights on all objects.  I am investigating these and will update things when they are clear.  Solution to this is here.

Each node can support 4 concurrent users, up to a max of 8 nodes.  More information can be found here in a vC Ops sizing document.  Important info!

You may only be able to log in as user@domain, or domain\user and more info can be found here on why.

In the vSphere Web Client, in the home screen, the vC Ops button will still be vC Ops button after your upgrade.  There is no workaround but almost no more info can be found here.

Summary

You should now have vR Ops 6 running in your environment with you vC Ops information!

I am not sure how I like this article in terms of how I pull other info in.  I like doing that for the background info, or the ‘extra’ info, and I like to show in my article the install / configure info.  This article was a bit of an experiment, meaning pulling in more info from others in the form of links – like for the install above. I think I will return next time to the way I have done it in the past – I detail the install / configure / upgrade stuff and reference additional materials elsewhere.  But we will see.

Thanks for reading, and if you have questions or comments let me know.

Updates

  • 7/7/15 – updated for 6.0.2 not having the log forward event issue mentioned above.
  • 3/12/15 – added link to article about how to share dashboards.
  • 2/15/15 – added KB article for login account info, and how 5.8 buttons in vSphere Web Client home screen remains after upgrade.  Also added link to Applications article.
  • 2/3/15 – small update for spelling and add up in pre-reqs info on mgmt pack compatibility.
  • 1/8/14 – updated with 4 concurrent users per node – max of 8 nodes.  Thanks William Lam!  Also added the info on how vR Ops log event sent via syslog can be seen with a – of hostname.
  • 1/6/15 – updated with the info that William Lam has done a script to make log forwarding much easier!
  • 1/6/15 – updated with the info to install third party MP in vR Ops before the migration to make sure third party data is migrated.

Michael

=== END ===

 

Tagged with: ,
Posted in How To

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: