Potential bug with Windows VM backup: "Body Timeout Error"

MajorP93

Any news on this?
If I recall correctly some work towards fixing the "compressed full backup of VMs with lots of free space" issue has been done but a fix was not (yet) pushed to the package repositories.

Is there a new test build that I could test or similar?
I would love to be able to use full backup method again.

Thanks!

olivierlambert

Let me ask internally

olivierlambert

This is the PR: https://github.com/xapi-project/xen-api/pull/6786

It's ready/reviewed, we are waiting for upstream merge. We'll make sure to re-ask for a merge ASAP.

Stay tuned!

last-genius opened this pull request in xapi-project/xen-api

closed stream_vdi: Only process allocated clusters for VHD and QCOW on XVA export #6786

MajorP93

@olivierlambert said in Potential bug with Windows VM backup: "Body Timeout Error":

This is the PR: https://github.com/xapi-project/xen-api/pull/6786

It's ready/reviewed, we are waiting for upstream merge. We'll make sure to re-ask for a merge ASAP.

Stay tuned!

Thank you very much for the update!

MajorP93

@olivierlambert @andriy.sultanov I saw that the linked PR got merged, which is awesome! I was wondering: do you plan to release this with the next round of patches or even before that as a hotfix?
In each case: if you have a test build I am happy to try it out.

Thank you and best regards

olivierlambert

Question for @stormi

MajorP93

@olivierlambert @stormi

Hey!
Any news in this regard?
Would be really interested in testing this fix in our lab environment if possible.
Thanks and best regards!

olivierlambert

Let me reping @stormi or @Team-XAPI-Network

psafont

@olivierlambert Because Andriy did quite a bit of work before this merge, xcp-ng depends on xenserver releasing the package to the clients before we can integrate it into xcp-ng. I'd say we can release the package February at the earliest

MajorP93

@psafont @olivierlambert As you previously mentioned february and we reached it as of now - I was wondering: do you have any update on when you plan to push the fix to XCP-ng repositories?

Thanks and best regards

psafont

@MajorP93 It's being packaged at the moment. between this and the testing that needs to be done it might take a couple of weeks to be released, but I can't promise anything yet

RS

I had same problem, about last 2 weeks, but is is about Xen update, after apply last update, Backup of Windows VMs works again.

gduperrey

The update mentioned by Pau should has just been released along with other updates. For more information, you can consult the corresponding blog post: https://xcp-ng.org/blog/2026/03/10/march-2026-maintenance-updates-for-xcp-ng-8-3-lts/

CodeMercenary

Amazing timing. I have been having this issue with a VM I just set up and I came here to find out if this is a known problem, only to discover that the 97 patches I have been planning to install tonight are going to fix it.

nikade

Did anyone try to patch xo and xcp-ng to the latest and did you have any luck with backups since then?
Im on XO (Xen Orchestra, commit 449e7 and Master, commit 2aff8) and the latest xcp-ng 8.3 with all patches installed and im still seeing the "Body timeout error" on 4/8 vm's:

Some works tho:

Observium has a 20Gb VDI with 15Gb used, jupiter has a 20Gb VDI with 12Gb used.
Let me know if you want any logs from my XO or XCP and I'll happily provide them.

CodeMercenary

@nikade I did install the 97 XCP patches and that night the VM that was failing to backup was able to successfully backup. It had a 500GB secondary drive that was formatted but totally empty.

I had not yet patched my XO from source to the latest, it was maybe 24 commits behind. On Friday I updated it so I will find out on Monday if those backups continued to work but I assume they did.

nikade

@CodeMercenary alright, thanks for replying. I tried some more, but still getting the same error, on linux as well as windows vm's so im at lost here.

Pilow

@nikade Hey Nikade,

Did you try to create a new job that will do a new chain ? Just for test.

nikade

@Pilow We tried that as well, same problem

Also tried with a VM on the same newtork, just another VLAN, and we're seeing the same thing. At first we figured it was because one of the xcp-ng's was on a remote site which is connected through an IPSEC VPN, but that wasn't the case.