... from the AMI creation script:

# prepare directories to store data
sudo mkdir -p /opt/clearml/data/elastic_7
sudo mkdir -p /opt/clearml/data/redis
sudo mkdir -p /opt/clearml/data/mongo/db
sudo mkdir -p /opt/clearml/data/mongo/configdb
sudo mkdir -p /opt/clearml/logs
sudo mkdir -p /opt/clearml/config
sudo mkdir -p /opt/clearml/data/fileserver
sudo chown -R 1000:1000 /opt/clearml/data/elastic_7

So it seems the AMI is using the correct directories... Do you have these?

hhrrmm.. in the initial problem, you mentioned that the /var/lib/docker/overlay2 was growing large in size.. but.. 4GB seems "fine" for docker images.. I wonder .. does your nvme0n1p1 ever report like 85% or 90% used or do you think that the 4GB is a lot ? when you restart the server, does the % used noticeably drop ? that would suggest tmp files inside the docker image itself which.. is possible with docker (weird but, possible)

After making the change yesterday to the docker-compose file, the server is completely unusable - this is all I see for the /dashboard screen

we turn off the server every evening...

In that case the issue is definitely not related to the mount points

it looks like clearml-apiserver and clearml-fileserver are continually restarting

no, they are still rebooting. i've looked in /opt/clearml/logs/apiserver.log no errors

yeah, that's usually the case when you get an empty dashboard

Try looking at their logs

strange, I used one of the publicly available AMIs for ClearML (we did not upgrade from the Trains AMI as started fresh)

🤔 i'll add the logging max_size now and monitor over the next week

thanks @<1523715084633772032:profile|AlertBlackbird30> this is really informative. Nothing seems to be particularly out of the ordinary though

3.7G	/var/lib/
3.7G	/var/lib/docker
3.0G	/var/lib/docker/overlay2

followed by a whole load of files that are a few hundred KBs in size, nothing huge though

Oh, that's strange. I'll run one of those soon to see if there's anything wrong with them

Morning, we got to 100% used which is what triggered this investigation. When we initially looked at overlay2 it was using 8GB, so now seems to be acceptable.

Howdy and Morning @<1687643893996195840:profile|RoundCat60> .. docker when using overlay2 doesn't have it's mount points show up in a 'df' btw, they will only appear in a 'df -a', mostly because since they are simply 'overlays', they don't (technically) consume any space (I mean, the files are still in the /var/lib but not for the space counting practices used by df)

this is why I was suggesting a find, maybe with a 'du' .. actually.. let me try that here.. 2s

back up and running again, thanks for your help

Hey there waves

Not sure about plans to automate this in the future, as this is more how docker behaves and not really clearml, especially with the overlay2 filesystem. The biggest offender usually is your json logfiles. have a look in /var/lib/docker/containers/ for *.log

assuming this IS the case, you can tell docker to only log upto a max-size .. I have mine set to 100m or some such

container_name:
  logging:
    options:
      max-size: 10m

Basically whatever was under the old /opt/trains/ folder is required, you can see the list here: None

also, is there a list anywhere with the mount points that are needed?

@<1687643893996195840:profile|RoundCat60> you set it once, inside the docker-compose itself.. it will affect all docker containers but, to be honest, docker tends to log everything

@<1687643893996195840:profile|RoundCat60> can you verify all the volume mounts point to existing directories on the server machine? (i.e. /opt/clearml/... )

thanks Stef, with max-size do you set it for every running service separately, or can you set it once?

not entirely sure on this as we used the custom AMI solution, is there any documentation on it?

I believe you can set it on a 'per container' way as well.

Hi @<1687643893996195840:profile|RoundCat60> ,
We've actually never had to address this issue. Can you find out what exactly is growing in size? I'd like to make sure this is not due to the containers storing data internally (causing docker to store more and more snapshots) - this is an unhealthy situation that might also indicate that volumes are not mounted correctly (i.e. data that should be stored externally is actually stored internally)

so am I right in thinking it's just the mount points that are missing?based on the output of df above

not yet, going to try and fix it today.

if I do a df I see this, which is concerning:

Filesystem      Size  Used Avail Use% Mounted on
devtmpfs        3.9G     0  3.9G   0% /dev
tmpfs           3.9G     0  3.9G   0% /dev/shm
tmpfs           3.9G  928K  3.9G   1% /run
tmpfs           3.9G     0  3.9G   0% /sys/fs/cgroup
/dev/nvme0n1p1   20G  7.9G   13G  40% /
tmpfs           790M     0  790M   0% /run/user/1000

so it looks like the mount points are not created. When do these get created? I thought using an AMI these would have already been setup?

Not necessarily, is there any data in those directories?

yep, in most of them:

/opt/clearml/config
apiserver.conf
clearml.conf

/opt/clearml/data/elastic_7
/nodes

/opt/clearml/data/fileserver
<empty>

/opt/clearml/data/mongo/configdb
<empty>

/opt/clearml/data/mongo/db
collection/index files, /diagnostic.data, /journal etc

/opt/clearml/data/redis 
dump.rdb

/opt/clearml/logs
apiserver.log.x, filserver.log (0 bytes)

incidentally we turn off the server every evening as it's not used overnight, we've not faced issues with it starting up in the morning or noticed any data loss