Hi, I am currently deploying Datahub using the v0....
# troubleshoot
m
Hi, I am currently deploying Datahub using the v0.8.40. My problem is that, the first time Datahub starts it has to create a lot of new indices in ElasticSearch. As this takes some time, in the middle of the process I get the
Copy code
Received signal: terminated
Command exited with error: exit status 143
Is there a way to modify the time that this graceful termination is executed??
s
Where are you deploying DataHub? How long is your GMS is taking to start?
m
I am deploying Datahub on Docker. It takes around 6 minutes.
s
Can you post the output of
docker ps -a
? Does
elasticsearch-setup
complete successfully?
How much resources is docker assigned? You can find via the Settings option of Docker Desktop
e.g. this is on my laptop
m
Yes, all dependencies are up and working perfectly. The status on ES is green and all de nodes are up as expected.
I am not using Docker Desktop, I am using instead Portainer with Docker Swarm
s
not really familiar with Portainer with Docker Swarm
you will have to check why the orchestrator is killing gms
I am not sure whether they have some liveness check or not
m
Yeah my docker compose looks pretty much like the one of the quickstart, I used it as a reference
So Datahub doesn't have internally any type of timer that if it reaches its limit the service is stopped?? It should be a timer of Docker ??
s
GMS getting killed seems to be done by the orchestrator. Have not really worked with Portainer with Docker Swarm. You will need to read their docs to see what liveness/readiness settings it uses.
m
Okey, thank you very much Aseem!
s
Actually there seems to be something https://github.com/datahub-project/datahub/blob/master/docker/datahub-gms/Dockerfile#L58 but have not really seen it getting killed
m
But as the request to the URL will probably result in a response of
404: Page Not Found
as the gms servive it is not still up. It could end up ending (although not gracefully)