Connection Server - Patching and Rebooting

[pbastiaans]

Looking for suggestions on how people are applying patches...actually handling the reboot post patch with multiple connection servers. They can be temperamental if not rebooted in order.

We have heard differing methods for handling reboot:

1. Shut them all down, reboot 1st, wait for it to come all the way up...reboot 2nd, rinse, repeat.
2. Shut them all down, power them all up at the same time

Our patch schedule is 1 connection server is considered 'DEV' and gets patched(and rebooted) a week prior to the other two, considered 'PROD'.

We recently ran into these servers having sync errors. We were scratching our heads trying to figure out 'what changed' and the subject of patching (the reboot) came up.

How are you handling the 'care and feeding' of these temperamental systems?

Thank you and drive fast, take chances.

[BenFB]

I patch my connection servers together to ensure consistency. First the dev/test environment to allow for testing. Anywhere from a few days to a week later I patch prod.

After each reboot run the following command from each connection server to verify replication is working.

C:\Windows\System32\repadmin.exe /showrepl localhost:389 DC=vdi,DC=vmware,DC=int

[pbastiaans · Accepted solution]

Through experience we watch the Horizon Console dashboard on all of our connection servers, if they are all in sync, no action taken. This is our 'canary in a coal mine'.

[Olgir]

We do it like this: patch one, reboot it, check if there are any replication issues or other kind of problems. If everything is okay, patch next one, reboot, check... etc.