RNHTTR commented on issue #26542:
URL: https://github.com/apache/airflow/issues/26542#issuecomment-1816984824

   > Yes. You need to run similar monitoring and restarting of Celery on your 
own. Liveness Probe runs periodicaly check if any component is responding and 
will restart it when it is not. This is standard feature of K8S and since you 
chose tto manage airflow on your own you should add similar monitoring on your 
own. You can take a look at the PR and take it as an inspiration when writing 
yours.
   > 
   > Generally speaking it's a good practice to do the kind of monitoring and 
liveness probes, and if you choose to run your own deployment, you shoud make 
sure you do similar check - this is standard practice with any kinds of 
applications.
   > 
   > Our Helm chart provides all those things out-of-the-box as an inspiration 
for anyone who woudl like to continue using their own deployment (you can 
easily run `helm install --dry-run` and you will see K8S resourrces generated, 
if you look for liveness provbes in varioous components you might see what CLI 
commands you can use.
   
   I actually don't think the livenessProbe solves the issue in all cases. It's 
rare, but I still see this issue cause workers to become useless when they lose 
connection with Redis


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to