3.4 RHEV-M should be on RHEL 6.5, and we tell users to use RHEL 6.6 for 3.5 RHEV-M. Our current upgrade procedure only provides instructions for upgrading from 3.4 to 3.5, but does not mention the underlying OS upgrade.
SME contact: Nikolai Sednev
QE team's upgrade workflow:
First things first, we did 3.3->3.4 as follows:
Setup with all hosts at 3.3 and engine 3.3.
Upgraded only the engine 3.3->3.4.
Upgraded hosts 3.3->3.4
Changed host clusters 3.3->3.4
Changed DC 3.3->3.4
Then we performed 3.4->3.5 as shown bellow:
Upgraded several hosts 3.4->3.5
Upgraded engine only first to 3.5 from 3.4
Upgraded engine's OS from 6.5->6.6.
Upgraded several host clusters 3.4->3.3
We still have to complete DCs upgrades 3.4->3.5
I'll give my comments regarding chapter 5 only, as it deals with RHEVM upgrade procedure, which we actually performed starting from 3.3->3.4->3.5.
5.1-1- We actually performed "yum update all -y" instead of updating only rhevm package, that is for two reasons, to not to miss kernel updates and to get OS updated from RHEL6.5 to 6.6 using correct repos.
5.3- We used RHEVHs and RHELS, during upgrade from 3.3->3.4 we decided to upgrade first the engine, after engine was upgraded we set hosts to maintenance, then upgraded them using engine, then rebooted manually because some mount bugs we had with NFS, we had to reboot them manually, otherwise host's could not mount their paths correctly.
I personally don't see anything more to be added to existing document, yet we didn't followed it during our upgrade task as we didn't knew of it's existence at a time of upgrade was done, we also decided to make upgrade procedure as simple as we could get and as close to our day-to-day path of upgrades.
Hi Julie, Lucinda, Nikolai
There is nothing additional to add from the network side regarding the upgrade flow and process.
The upgrade procedure and process well covered.
3.4 RHEV-M should be on RHEL 6.5, and we tell users to use RHEL 6.6 for 3.5 RHEV-M. Our current upgrade procedure only provides instructions for upgrading from 3.4 to 3.5, but does not mention the underlying OS upgrade. SME contact: Nikolai Sednev QE team's upgrade workflow: First things first, we did 3.3->3.4 as follows: Setup with all hosts at 3.3 and engine 3.3. Upgraded only the engine 3.3->3.4. Upgraded hosts 3.3->3.4 Changed host clusters 3.3->3.4 Changed DC 3.3->3.4 Then we performed 3.4->3.5 as shown bellow: Upgraded several hosts 3.4->3.5 Upgraded engine only first to 3.5 from 3.4 Upgraded engine's OS from 6.5->6.6. Upgraded several host clusters 3.4->3.3 We still have to complete DCs upgrades 3.4->3.5