I want to say this is likely normal depending on the SAN you are using. I have worked with dozens of vendors over the years. Some allow ESXi to boot and mount the iSCSI very fast while others seem to take a few extra minutes.
I am currently managing Dell EMC scv3020 arrays and there is about a 1-2 minute delay during boot when it is bringing up the iSCSI. I have 2 sites doing the same thing. One is presenting 40 LUNs the other only 5 and the delay is equal.
I never really dug too deep into it because in the end it just works once the OS has loaded. I have a feeling (at least in my situation) that it is trying to mount something from the passive controller and eventually timing out. VeeamOne send me an "errror" that it could not initialize one path. EMC told me this was expected behaviour based on how we were configured.
When I managed 3PAR or Hitachi storage, the volumes would mount almost instantly.
What array type did you migrate to? In your situation, 35 minutes is excessive and likely something is not properly configured. Have you worked with the storage vendor to diagnose this?
Example of what I see :
Alarm: | Connection to iSCSI storage target failure |
Status: | |
Previous status: | Reset/resolved |
Time: | 7/11/2023 1:42:22 PM |
Details: | Fired by event: esx.problem.storage.iscsi.target.connect.error
Event description: Login to iSCSI target iqn.2002-03.com.compellent.somenumberstring on vmhba66 @ vmk2 failed. The iSCSI initiator could not establish a network connection to the target.
Initiated by: Not Set |
