[KB2136854] There's a new CBT bug in ESXi 6

[Gostev]

"Virtual Machine Migration or Shutdown Required: Yes"

But this does not say a word about having to reboot your VMs ?!!!

[cffit]

I guess to me it reads that you have to either do a migration of the VM (not sure if that means storage migration or host migration) or shutdown the VM. And "shutdown required" makes me think the VM must be rebooted. How does that read to you?

[bgbGsy]

Hi all.

Obviously, the last few months have seen a number of CBT related issues with both 5.5 & 6 Esxi. This last one was particualrly worrying as it could have been potenitally happening for some time.

One suggestion for a future Veeam release might be an option to 'reset CBT on next Run' for resetting the CBT. I imagine Veeam has the logic for this now due to the issue with extending disks. Ideally the option could be selected and then it would automatically return to the 'off' position after a sucessful backup.

Of course PowerCLI can be used, but this would be an elegant solution.

Thanks.

[cffit]

I guess to me it reads that you have to either do a migration of the VM (not sure if that means storage migration or host migration) or shutdown the VM. And "shutdown required" makes me think the VM must be rebooted. How does that read to you?

I think I realize what "shutdown required" means now. I was thinking the way the article read that you had to patch and reboot the ESXi host for the patch, and then after that you had to migrate or shutdown the VMs as well afterwards. What the article probably mean is you have to migrate or shutdown the VMs so that you can reboot the ESXi host. My bad in interpreting it the wrong way.

[lando_uk]

Hi,

So after all this talk, is ESXi600-201511001 3247720 now good for production?

We have new hosts and a task to update to the latest stable version, is everyone now confident on 6, or are people still rolling out 5.5 just to be safe?

Thanks.

[Gostev]

Are you willing to risk your Christmas holidays to find out if it's good for production?

Remember - it's a hot fix that had received only a few days of testing.
I personally would wait for ESXi 6.0 Update 1b before upgrading...

[mdking]

Gostev,

Do we believe that the patch is working correctly? A few pages back the patch was tested in the same environment and seem to be working fine. Now you mention that the patch might not be?

Thanks
Matt

[lando_uk]

Are you willing to risk your Christmas holidays to find out if it's good for production?

Remember - it's a hot fix that had received only a few days of testing.
I personally would wait for ESXi 6.0 Update 1b before upgrading...

Hi Gostev

Well, I'm not starting any upgrades until the new year, its just the new servers have been delivered and I need to at least decide on a plan.

A rubber stamp from you guys would give us confidence. Update 1b sounds good to me... I don't have to start anything until Feb really.

Thanks

[lightsout]

Patch seems to be working fine to me!

[Gostev]

Now you mention that the patch might not be?

No-no, don't get me wrong here. All I am saying is that it is too early to tell with confidence.

[edvschriesheim]

Am I right?

If one of my VMs are affected by this bug Veeam B&R8 would warn me?
According to KB1940 Veeam can detect corrupt CBT Data.

[veremin]

Nope, as the referenced KB talks about different CBT issue.

[Pat490]

VMware just released ESXi 6.0 update 1B!
http://pubs.vmware.com/Release_Notes/en ... notes.html

CBT bug is listed as a resolved issue. I hope that there will not be any new bugs in that release!

[veremin]

It seems so, indeed. But just as is with every other updates be sure to take safety measures - back VMs up prior to upgrading, and watch closely how the backup and replication activities go after the upgrade (surebackup and surereplica functionalities should help you with that). Thanks.

[lando_uk]

Or just wait a month and let some other mug be the guinea pig

Looking at date and build of update 1 (2015-09-10, 3029758), I would have thought we'd be due an Update 2 rather than a `b`

[veremin]

Agreed. Given the number of bugs discovered with 6 release, waiting would not hurt, indeed.

[mcvosi]

Once 1b is applied we'll need to reset CBT again?

[Gostev]

Yes, unless you did that already after installing the earlier VMware patch for this CBT issue.

[mcvosi]

OK thx

[jtek]

I didn't see a response to this. Does this mean it worked because I was getting the same response running the mass update with "Warning: Parameter 'VM' is obsolete. this parameter no longer accepts multiple values."
Thanks!

Thanks Gostev, didn't see the link, I ran the ResetCBT.ps1 and it seemed to work but I did get a wanrning in between creating and removing snapshots (it did create) I presume this isn't a problem ?

[jtek]

Just wanted to reply to my last post. I ran this on PowerCLI 6 release 3 and got the VM parameter issue. Just ran it again on my laptop with PowerCLI 6 Release 1 and didn't generate the parameter message. It ran as expected.

Thanks, J

[rkoppelman]

To my knowledge vSphere 6 patch 1b solves the issue of this toppic so this threat should be closed

[Gostev]

And so it is