CDP with CBT mode - Warning

[jmc]

Hi all,

after creating a cdp job with 3 vm's i get the message CBT Mode instead of syncing on one vm. This results in the warning:

Restore point creation completed with warnings. (Failed to create restore point 05.04.2021 12:00:00 (attempt 1 out of 3)) Details: CBT mode was enabled for the following disks: Hard disk 1

what does this mean and how can it be cleaned?

thx
jeff

[nikolaj]

Hi jmc,

CBT is a vmware technology that is utilised by VBR to create incremental backups.
The setting whether or not to use it can be controlled in the Advanced settings of VBR console:
https://helpcenter.veeam.com/docs/backu ... ml?ver=110

In your specific case it looks like there's a mismatch between this setting in VBR and the configuration parameters of a particular HDD.
Is CBT was indeed enabled for Hard disk 1 in vSphere?
https://kb.vmware.com/s/article/1020128

What are the settings for CBT in Advanced settings > vSphere of your VBR instance?

[jmc]

hello nikolaj,

you are right and i know this. but that isn't what i asked. WHY i get for one VM in a CDP job CBT mode and the other vm's are syncing? in all b&r jobs the cbt is activated also for the correct vm's with syncing.

jeff

[veremin]

In case of CDP, CBT mode does not refer to VMware Changed Block Tracking mechanism, it describes a case when a source proxy cannot deliver the traffic stream to target infrastructure and instead of sending blocks it starts creating a changed block map (identifying those blocks that have changed during the time period when there has been no connection to the target infrastructure).

When connection goes up, CDP sends to the target the actual content of the source blocks - those blocks that have been marked as changed ones in the block map.

Thanks!

[jmc]

hi vlademir,

thx for explanation. what i couldn't understand is, why this happened with this specific vm. the other vm's are in syncing mode. the ressourcepool, the vm-cluster and the hosts are equal. i got the error:

Code: Select all

06.04.2021 08:00:17 :: Using source proxy 192.168.65.63 for disk Hard disk 1  
06.04.2021 08:00:17 :: Using target proxy 192.168.65.63 for disk Hard disk 1  
06.04.2021 08:00:17 :: Processing disk: Hard disk 1  
06.04.2021 12:00:00 :: Restore point creation completed with warnings. (Failed to create restore point 06.04.2021 12:00:00 (attempt 1 out of 3)) Details: CBT mode was enabled for the following disks: Hard disk 1 CBT mode was enabled for the following disks: Hard disk 1 
06.04.2021 12:00:00 :: Creating crash-consistent long-term restore point at 06.04.2021 12:00:00  
06.04.2021 12:10:00 :: Restore point creation completed with warnings. (Failed to create restore point 06.04.2021 12:10:01 (attempt 2 out of 3)) Details: CBT mode was enabled for the following disks: Hard disk 1 CBT mode was enabled for the following disks: Hard disk 1 
06.04.2021 12:10:00 :: Creating crash-consistent long-term restore point at 06.04.2021 12:10:00  
06.04.2021 12:20:01 :: Restore point creation completed with errors. (Failed to create restore point 06.04.2021 12:20:01 (attempt 3 out of 3)) Error: CBT mode was enabled for the following disks: Hard disk 1  
06.04.2021 08:00:17 :: Processing disks...  

any hint or should i open a ticket?

thx jeff

[veremin]

Hard to guess really, so opening a support ticket sounds like the best idea. Not only will it allow us see the debug logs and your infrastructure, but also it will help me to draw attention of QA team to the issue reported. Thanks, Jeff!

[mganadi]

Hi,
Is there any update for this problem I have the same my CDP is in cbt mode !!

[Gostev]

This is normal in case of an insufficient bandwidth. However, if you don't see your available bandwidth fully utilized, then you should open a support case for further troubleshooting. Thanks!

[jmc]

hi,

bandwidth is not the problem. i ahbe about 30 vm's also very large and a 20 gb line. all run normally. only this one of 5 gb does not come out of the cbt.

thx
jeff

[veremin]

Then please share with us a support ticket number, so we can bring the issue internally. Thanks!

[bhagen]

Any updates on this subject? I'm experiencing the same thing now and submitted a ticket: Production Support — ID: #02034922

[veremin]

I have checked the case and it looks like some information is still missing: logs from CDP policy, hosts, clusters, vCenter server, etc. Can you please collect those and attach them to the existing ticket? Thanks!

[bhagen]

Gathering them now, thanks!

[veremin]

Let us know when they are uploaded, so we can ask QA team to the check the case. Thanks!

[bhagen]

Uploaded...finally...

[veremin]

Thanks! I have shared this information internally. I will update the topic, once I have more information. Thanks!

[veremin]

The case is still missing most of the required logs, so we will proceed with the investigation, once those are provided. Thanks!

[bhagen]

Logs are now uploaded.

[veremin]

The issue has been confirmed, it is fixed in the upcoming product version - 11a. Thanks!

[bhagen]

It's a bit sad that I had to read this on the forum rather than hearing it in my support ticket.

But at least it's being addressed.

In the meantime, I'll disable all my CDP policies, since they're all exhibiting this issue.

[veremin]

They would come to the same conclusion, we just have a benefit of immediate access to the brains of our QA team (which I've asked to check the case), so some things are clarified bit faster this way.

By the way, 11a release should happen next month.

Thanks!

[DTB]

Hi...

I've this issue after patching V11a to 11.0.1.1261 P20220302

Is there a solution???

[andre.simard]

Same issue also.

We have two VM in the CDP policy. One in sync mode and the other one in CBT mode

i'm on 11.0.1.1261 P20220302

Case # 05379976

[veremin]

I see that the investigation still goes on, so let's wait till it's finished and make decisions based on the final outcome. Thanks!

[andre.simard]

Hi,

I got an answer from the support that Veeam backup on a vm that has CDP job is not supported. is it normal?
In the documentation it says that after the initial replication snapshot created by veeam replication are supported.

VMs that you plan to protect must not have snapshots at the moment when the CDP policy starts for the first time. Further, you can add these VMs to backup jobs that will create snapshots. However, you must not create snapshots by other means, for example, manually or by another software. This may disrupt further replication operations or cause loss of data.

https://helpcenter.veeam.com/docs/backu ... ml?ver=110

[veremin]

The answer provided by the support engineer was a bit different. She suggested that CDP policy is always switched to CBT mode, when there is an active VM snapshot created by backup job, and gets back to processing, once the snapshot is deleted. This is partially correct: CDP policy might be switched temporarily to CBT mode, when backup job creates a VM snapshot - but it should not last longer than a few seconds (definitely not till the backup job ends).

We have reached your support engineer to clarify the details and provide assistance with the investigation.

Thanks!

[veremin]

R&D team have confirmed that you faced the specific bug. Currently the team is working on a private fix for it. The support engineer will share the fix as soon as it's created. Thanks!

[veremin]

The private fix has been created and certified (every time the code interacting with certain VAIO components is updated, the whole bundle needs to be recertified by VMware). Soon it will be delivered to you. Thanks!

[andre.simard]

Thank you for the update veremin!

[veremin]

It was shared with the support team today. The support engineer should provide you with it shortly. Thanks!