I have modified sample DIY Backup scripts to work with Microsoft Azure Backup Service (MARS). The issue I am getting is that when MARS is running a backup (while Bictbucket is in maintenance mode) it encounters issues with gc.log and sometimes peeled-tags files across various repositories.
These files are accessible well after a backup is complete and they are reported as 0kb, however, when I open a file with a text editor I can see some contents (few lines) and then kb size is updated since a flush is triggered.
At the end of the backup MARS reports a warning and logs a list of files with DATA FAILURE 0x80070002, most of which are due to gc.log.
Since Bitbucket 5+ has implemented its own garbage collection as per this article (https://confluence.atlassian.com/bitbucketserverkb/how-to-perform-a-manual-garbage-collection-on-a-repository-867194296.html), I am hoping someone can shed some light into this issue and how to avoid it.
We are running Bitbucket Server v5.16.0.
Atlassian does not support modified versions of the script, and we can't guarantee that will work, so anything we do now will be on a best effort basis.
To better understand why are you getting those errors, can you share what kind of modifications you made to the script? Also, what are the issues reported by the tool?
Let us know!
I was replying to you and managed to reply to myself... re-posting my reply here...
I am not sure how this was marked as an accepted answer, I guess I might have clicked it by accident.
Whatever I have posted in my previous post was a clarification to my original post and an attempt to answer the questions you have raised.
Suggested UI improvement:
Your color coding scheme for the Accept/Accepted answer buttons is counter-intuitive. Granted, I should read what it says but a green button with a checkbox should mean the answer was accepted and a white button with green border should prompt user to accept the answer.
Can anyone reply to let me know this question is still being looked at?
I am actually not looking for help with the script itself. It is more to do with an access to gc.log files. Once I place Bitbucket into maintenance mode, as per your standard script, I activate and run Microsoft Azure Backup which performs a DB backup as well as an incremental backup of Stash Home directory. Here are the steps from Microsoft site describing the backup process.
So, when step 3 is executed (snapshot is taken) the process encounters warnings when accessing gc.log files. Since gc.log files are manged by Bitbucket (since v5) I am reaching out to Atlassian to see what can be done here. Is there some maintenance to be performed on the affected repos?
Did you ever get an answer to this @Vitalii Djiguir ?
@Ana Retamal We are also experiencing Azure file backups completing with errors - those errors being some repositories gc.log files under home\shared\data\repositories\#\app-info\gc.log -
1) Can this be safely ignored ?
2) Can I delete these files ?
3) Can i tell the backup not to try to backup gc.log files ?
4) Are these files important ?
Hi everyone, We are looking to learn more about development teams’ workflows and pain points, especially around DevOps, integrations, administration, scale, security, and the related challeng...
Connect with like-minded Atlassian users at free events near you!Find an event
Connect with like-minded Atlassian users at free events near you!
Unfortunately there are no Community Events near you at the moment.Host an event
You're one step closer to meeting fellow Atlassian users at your local event. Learn more about Community Events