Patching for VBO infra best practices and issues

[rtheseeker]

We have one VBO server and 7 proxies and all sit in Azure; at the moment we patch these proxies (4 on one day; 3 next day plus VBO); is there any best practices around this; we only reboot the proxies/VBO that we are patching.

Other issue that I have noticed is that the backup copies go extremely slow after the proxy reboot; will patching/rebooting the whole envrioment would be better approach and is it expected for the backup copies to be slow after the proxy reboot. I already have a case open with Veeam and it is with 2nd Level however I did not get any definative answer as to if this is a expected behaviour (Veeam Backup for Microsoft 365 #03110346)

I can definetly see huge delays in terms of backup copies where the job which usually takes less than a minute to process a mailbox is usually sitting there for over 20 minutes everytime we reboot a proxy (no previous jobs running); it comes right after the first job finishes on most ocassions; however (senario 2)randomly the backup copy goes slow.

I had been advised to update some dlls which I had done already and was given another set; but I held that as thought of first upgrading to the most recent version of VBO (P20240123 (7.1.0.1501))

So the question is
a. Is backup copy slowness expected after rebooting a proxy; it copies data from Azure Blob to Aws; this reboot does not impact backup
b. Is there any best practices related to patching VBO envrioment with multiple proxies in diffrent AZ

[Mike Resseler]

a) No, this is not expected. After a reboot of a proxy it should run as normal. However, we do see that sometimes a job is slower when the proxy was rebooted while the backup job was running. Was this the case?
b) v8 will make a huge difference in this as you will be able to put proxies in maintenance mode (you will be able to create a pool with those 7 proxies and they will be able to divide the backup load amongst them... then you can put one in maintenance mode, it will flush out its current work to other proxies and you can do windows update or something like that)

[rtheseeker]

Thanks I will ask support to investigate further; now all the backup copy jobs are very slow; for example the mailbox was taking 11 minutes to copy now has completed in 47 minutes; so we have 4997 objects in an exchange backup job (we have multiple exchange jobs across multiple proxies) so you can understand as it would be days before the this is copied to AWS objectstore;

Also just to double check Is there any best practices related to patching VBO envrioment with multiple proxies; should the whole envrioment be patched togeather or we can continue to patch the way we are (4 proxeis first and rest 3 and VBO the next day)