Open Ashkaan opened 1 year ago
Good news: I tested a few versions, and the last version that was working is: semaphoreui/semaphore:v2.9.18
I tried v.2.9.21, v.2.9.23, whatever latest points to, and whatever develop points to, and none of those work.
i'm experiencing the same issue. After a rollback to version v2.9.18 everything works again.
It seems that all versions above 2.9.4 are NOT considered as stable but as pre-releases. Not sure why there is no stable docker tag.
Same issue: #1485
Consider to use v2.9.18
Fix will be soon.
Please check latest pre-release (latest
docker image).
Did not work for me. New issue: SSH key not working. I put it in and overwrote, now jobs are stuck on "Running".
@Ashkaan need logs.
@Ashkaan are you have custom config.json
or passing settings via environment vars?
I had to revert for production.
Why I asked.
Container writes config.json
only if it doesn't exist. So, if you stored config.json
in volume and recreate container with different env vars, it will not affect to config.json
stored in volume.
Currently this behaviour changed because we maps env vars to config fields. Env vars have priority over config.json
.
Oh interesting. Yes, that might be the case. The config.json is mapped externally.
@Ashkaan thank you very much! I will think what to do... I didn't expect these to be breaking changes.
@Ashkaan could you remove environment variables from docker image if you use mapped config.json
?
It doesn't load at all without env vars.
@Ashkaan got it, thank you.
Please try latest pre-release: https://github.com/ansible-semaphore/semaphore/releases/tag/v2.9.36
It just hangs. I tried to go back to 18, and it wouldn't run also, so I had to restore the database from backup.
Hi @Ashkaan Cloud you run job in DEBUG mode?
I am facing the same problem, I think. It seems to be related to the ssh control path.
If I manually enter the command it is stuck on in the console, and delete the ControlPath, it works.
2.9.39-beta
I am facing the same problem, I think. It seems to be related to the ssh control path.
If I manually enter the command it is stuck on in the console, and delete the ControlPath, it works.
2.9.39-beta
Use version 2.9.37
I am facing the same problem, I think. It seems to be related to the ssh control path. If I manually enter the command it is stuck on in the console, and delete the ControlPath, it works. 2.9.39-beta
Use version 2.9.37
Downgrading to 2.9.37 did not resolve the issue.
11:17:30 AM
PLAY [all] *********************************************************************
11:17:30 AM
11:17:30 AM
TASK [Gathering Facts] *********************************************************
11:17:30 AM
task path: /var/lib/semaphore/tmp/repository_1_8/playbooks/netbox-agent/netbox-prod.yaml:2
11:17:30 AM
<10.1.90.118> ESTABLISH SSH CONNECTION FOR USER: lxadmin
11:17:30 AM
<10.1.90.118> SSH: EXEC sshpass -d9 ssh -vvv -C -o ControlMaster=auto -o ControlPersist=60s -o StrictHostKeyChecking=no -o 'User="lxadmin"' -o ConnectTimeout=10 -o ControlPath=/var/lib/semaphore/tmp/.ansible/cp/c3de483d93 10.1.90.118 '/bin/sh -c '"'"'echo ~lxadmin && sleep 0'"'"''
stuck here.
2.9.18 is the latest version that works great for me.
I just tested several of the lastest version including betas and they're still not working. Any word on when I can go back to "latest"?
Thanks so much
I can agree. I just installed the newest version of the release version and the screen is not refreshing. v2.9.37 is not working. If you refresh the screen it show the job progress but it is not auto refreshing. I'd go the beta, but I tried install and it and my keys stop working.
@dennisaucoin is this issue actual? What screen? Need more details. Thank you
I'd go the beta, but I tried install and it and my keys stop working.
Your keys always stop working with any version change. Just input them again.
@Ashkaan Keys encrypted in database.
It was a difficult choice between not encrypting data by default, or generating a new key when deploying the container without SEMAPHORE_ACCESS_KEY_ENCRYPTION
. We chose the second option.
So, you have 2 options:
1) Backup /etc/semaphore/config.json
from container to host and mount it to container.
2) Pass environment variable SEMAPHORE_ACCESS_KEY_ENCRYPTION
to container: https://docs.semui.co/administration-guide/security.
I've had a great and working Semaphore docker for some time now. Suddenly, it hangs here when I run any task:
If I hit Stop, nothing happens. If I hit Force Stop, it sets it to a Stopped status. I don't have any tasks running. Is there a way clear the queue, or something?
I'm running this on docker using the default Bolt DB and I can access the CLI if needed.