VUM upgrade fails - ramdisk (upgradescratch) is full

[andvm]

Hi,

I am trying to push an ESXi 6.7 upgrade via VUM but keep getting this log (and fails) in vmkernel and the scratch partition is on a SAN datastore with plenty of free space

MemSchedAdmit: 471: Admission failure in path: upgradescratch/upgradescratch

MemSchedAdmit: 489: upgradescratch (289688) extraMin/extraFromParent: 1/1, upgradescratch (289687) childEmin/eMinLimit: 92160/92160

WARNING: VisorFSRam: 206: Cannot extend visorfs file /upgrade_scratch/QEDENTV.V00 because its ramdisk (upgradescratch) is full.

Any suggestions as to which disk/path it is referring to being full?

Thanks

[nirmalgnair]

Hi @andvm

This is a know issue and at the moment there is no work around.

You need to use offline bundle and upgrade the ESXi using CLI.

VMware Knowledge Base

Regards,

Nirmal Nair

[andvm]

Would it work if I reboot and boot directly from ISO image and select upgrade?

Does this bug belong to a specific vCenter or ESXi version?

Thanks

[nirmalgnair · Accepted solution]

Hi @andvm,

Booting from ISO will work but offline bundle will be easy if you have multiple hosts. You can put the offline bundle in a shared storage and execute the command to upgrade.

Also I prefer CLI as you just have to reboot the host only once.

This issue is only seen with VSCA 6.5/6.7.

Regardsm

Nirmal Nair

Did you find this helpful? Let us know by completing this survey (takes 1 minute!)

[andvm]

Ah ok will do that way as yes, have multiple hosts to upgrade.

Am listing steps below post placing host in MM.

• Download offline bundle and place on Shared storage. (Done only once)

• Run esxcli software vib update -d "/vmfs/volumes/Datastore/DirectoryName/PatchName.zip". (PatchName will be the ESXi offline bundle zip file)

• Reboot manually (think it will not reboot automatically) and verify ESXi version/build is as expected.

[nirmalgnair]

Yes, the steps are correct.

After the command is run, you will see a message that Reboot Is Required.

Regards,

Nirmal Nair

[amolnjadhav]

Even i have observed similar error while upgrading to 6.5U3 from 65U2.

@nirmalgnair, Does this issue exist in 6.5 as well?