ESXi 7.0U3g

[ddayton]

Two questions:
1. Is ESXi 7.0U3g fully supported with Veeam B&R 11.0a?
2. Is ESXi 7.0U3g have any known problems to watch out for?

I have a few standalone vSphere ESXi hosts running V6.7U3 and we are coming up on a multi week production outage, so a good time for upgrades. The environment here is very simple with no vCenter or any other extras. Hosts are all using local disks on LSI SAS RAID controllers.

[Andreas Neufert]

Minor Releases such as here the “g” one are automatically supported and do not contain usually any features or API changes, only bug fixes. So in the past they do not caused any issues for us.

Please make sure that you run the latest Veeam Backup & Replication v11a (11.0.1.1261 P20220302)

[ddayton]

Thank you. Veeam Backup & Replication v11a (11.0.1.1261 P20220302) is the version that I am on. Our Veeam server is on a VM, so I typically take a snapshot and install the Veeam updates soon after they are released. Once I check that my daily jobs have ran fine once then I delete the snapshot. Only one upgrade has ever required a call to support and that was a SQL Express problem on a major release. Veeam software and support are great!

[Andreas Neufert]

Thanks for the feedback.

[MaartenA]

People using iscsi Nimble Snapshots(and maybe other Vandoren)and have the ACL set on the volume to volume & snapshot (default setting) should be aware that the last esxi version will break your platform because of the stale iscsi paths created. Somehow the behavior is different compared to like the 7.0U3D release and during backup the host will hang shortly causing ping loss to the VM. We faced this issue and needed to revert esxi back to change the ACL on the volumes first to Volume Only.

[Andreas Neufert]

This is not a specific ESXi 7.0U3g thing.
Whenever you perform a storage rescan on the host cards within the ESXi, it can hang if there are snapshots present.
Basically the ESXi host will see the same volume ID multiple times (volume and snaoshots) and try to run something on a read only snapshot which will fail.
You are just lucky that you did not had any issues.

This issue is well documented within the Nimble + VMware best practices => To set the policy to "volume only".

[MaartenA]

Yes we know now but never faced this issue in 7.0.3D and earlier releases.