-
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
[BUG] regression in 3005.x - salt hangs after highstate for windows minion #63061
Comments
Please fill in the |
ok, done. |
This appears to be the same issue as #62937, but you already have the update with PyZMQ 22.0.3. Has the minion service definitely been restarted after the upgrade? |
Yes, the minion service has definitely been restarted. Yes, I agree it looks or feels like the same, but I am told that the locally run And strangely I have to agree, as I have seen myself that the manual direct upgrading of ( as per salt-users group email thread here: https://groups.google.com/g/salt-users/c/O5mAujEzjmc/m/XvCw0T-QBgAJ ) |
@twangboy do you have any insights or suggestions on how I could debug this better to find the blocker of this highstate output? |
A similar stacktrace here, but with a syndic: #62577 (comment) |
@TheBigBear Do you happen to be on a Multi-master set up? I'm wondering about your |
@twangboy, no we have a 'simple' single master setup. |
Do you have an example of the states you are running? In #62550 it was exposed by anything that communicated with the master fileserver, for example So far I have been unable to reproduce it. |
@twangboyI finally did what I should have done at the beginning. ;-) I let it display my "top.sls" for the windows minion and then went through all of the 6 states one by one to see which one actually has any issues.
produces following error output
And the windows minion rendered salt.minion sls is:
and for reference IF I run the relevant state directly on the windows minion itself using
and the same without the 'test=true' shows the same basically ( I think):
I hope and trust this helps you reproduce this? I am a bit surprised by it trying to use linux rights and permissions mechanisms and linux users like user 'root' and group 'root' to set permissions on a bunch of windows minion installation files and directories? So @twangboyt is this actually two separate issues in the end?
|
These may be the issue:
|
OK, I have reverted back to Note: now it properly reports the user as 'system' under the windows minion. remember under the 3005.x it reported it as 'root' instead. Here is the output of
|
I would like to see |
@dafyddj your wish is my command ;-)
salt master versions:
salt windows minion versions:
and
Will that do? Or do I need to capture a higher log-level or diagnostic level output as well do you think? |
Hey @twangboy, yes, I think this is the likely culprit, salt should be able to handle a mis-firing state and not break under its output, I think. I am setting up a home/test/lab to test the regression of going from But my python |
Would you mind verifying this is still a problem with 3006.0? I'm wondering if it's unable to resolve a SID:
|
@TheBigBear Could you share the contents of |
@twangboy I am using the salt formula and a pillar in the middle so what I can supply is the So this has remained broken ever since I upgraded from 3004.2 to 3005.x and remained broken even now on 3006.1 Here is my rendered above
So here is a step by step re-run of how to duplicate the issue under 3006.1 on the linux master it looks as follows:
And now I am running those 4 states one by one ( spoiler - only the
So that ran just fine. But now comes the culprit.
And just to show the last 2 states, they also run without any issues.
and the last one is fine too
BUT a
And this salt job never gets any feedback at all. However on the windows minion side using First this is how the
|
I think a big part of this regression is the fact that the salt-formula is not being maintained very well. I had created a report in the salt-formula repo back at end of 22 and it is still broken today as per this issue here. saltstack-formulas/salt-formula#541 |
Those formulas are trying to set permissions for the user
The salt-formula needs to be updated to provide valid users on Windows. These may not have been applied in earlier versions of Salt and that is why you don't see the error in 3004.2. |
@twangboy how is this ‘completed’? |
No. They're completely independent. |
Description
Since the upgrade to the latest version 3005.x on the master the windows minions no longer send highstate output back to linux master.
Setup
upgrade master to 3005.x from 3004.2 and any previously working test only
salt mywinminion state.highstate test=true
or an actualsalt mywinminion1 state.highstate
times out and only results in a generic:But this job output never gets any data even hours after having been run.
Please be as specific as possible and give set-up details.
Steps to Reproduce the behavior
(Include debug logs if possible and relevant)
Expected behavior
A clear and concise description of what you expected to happen.
Screenshots
If applicable, add screenshots to help explain your problem.
Versions Report
salt --versions-report
(Provided by running salt --versions-report. Please also mention any differences in master/minion versions.)Additional context
Add any other context about the problem here.
The text was updated successfully, but these errors were encountered: