-
Notifications
You must be signed in to change notification settings - Fork 5.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Bad gitfs_remote breaks sls-files in subdirectories for state.(sls|highstate) #21021
Comments
Thanks for this excellent report @JPT580! I can totally see your point. I think what is happening here is that we didn't want to make a bad gitfs remote fatal and just log the error. But if it isn't fatal, then people aren't necessarily looking at the log. I can also see your reasoning about wanting the master to fail to start when the config is bad. I don't think there currently is a method to make that fatal. The only fatality that occurs now is if no fileserver backends are configured for the situation where a user has roots and git configured but git is broken, we still have roots available. Anyhow, this is a great point that will take some careful consideration. I've discussed this issue with @terminalmage, our main gitfs man. |
@JPT580 I'll discuss this with some of the other core devs and try to come up with a way to make invalid FS backend configuration fatal. I really like the idea. |
@JPT580 What do you think of this?
|
This should be better to debug for most people. |
@JPT580 no, that's exactly what is happening. hence the "exiting" message |
Sorry, i misread that. It's perfect! |
This causes the master to fail to start when there are fileserver config issues, leading people quicker to the log to investigate, where they will find a meaningful error message. This should it much quicker and easier for users to realize there are gitfs/hgfs/svnfs fileserver config problems. Fixes saltstack#21021.
Fix was merged. Closing. Thanks again for the suggestion! |
This causes the master to fail to start when there are fileserver config issues, leading people quicker to the log to investigate, where they will find a meaningful error message. This should it much quicker and easier for users to realize there are gitfs/hgfs/svnfs fileserver config problems. Fixes saltstack#21021.
This causes the master to fail to start when there are fileserver config issues, leading people quicker to the log to investigate, where they will find a meaningful error message. This should it much quicker and easier for users to realize there are gitfs/hgfs/svnfs fileserver config problems. Fixes saltstack#21021.
This causes the master to fail to start when there are fileserver config issues, leading people quicker to the log to investigate, where they will find a meaningful error message. This should it much quicker and easier for users to realize there are gitfs/hgfs/svnfs fileserver config problems. Fixes saltstack#21021.
Version
This may probably be obsolete with #6739 in mind.
Setup
salt-master
, add valid gitfs_remotes which exist in the filesystem.salt-master
State tree in
/srv/salt
Master configuration
/srv/git/vmware-tools-formula
did not exist in my filesystem.Expected result
Restarting the master fails / A warning like "Invalid gitfs_remote in config ... does not exist in filesystem" appears.
Observed result
Log excerpt from an affected minion:
Command used:
salt 'salt-minion-1*' state.sls 'vmware-uptodate'
The text was updated successfully, but these errors were encountered: