Reputation
Badges 1
19 × Eureka!Not that I know of... Is there an agent waiting on the queue? If so, can you try without an agent listening on the queue?
Can you try a different browser with private mode on?
I'll probably be able to better figure it out next week. If you have some new info or find out more I'd love to hear about it 😄
No, I was wrong... 😞 - the error is actually returned from the apiserver
itself, so that works...
Just want to make sure the agents are reporting but for some reason you're missing it 🙂
ColossalAnt7 can you reach the UI from your browser?
Since your aim is to end up with content you can paste into apiserver.conf
(which is not loaded by Trains)
However, that can easily be done using ConfigFactory.parse_file("apiserver.conf")
it will return a
Config
object right?
Will return a ConfigTree
object which behaves the same way
What about:apiserver_conf.put( "auth.fixed_users.users", list(apiserver_conf.get("auth.fixed_users.users", [])) + [{'username': username, 'password': password, 'name': name}] )
Hi SmarmySeaurchin8 ,
The agents that report to the server are registered in the Redis component and their registration is timed out after a period of 10 minutes (assuming they're not still up and reporting).
Is the agent in question still up? Also, how was the server reset?
@<1576381444509405184:profile|ManiacalLizard2> can you try putting it in the repository's requirements.txt instead of adding it in code?
so if you have very large snapshots that are close to one another one might wait for the other for quite some time
yeah, it is async, but when talking a snapshot it will wait for the previous model to finish uploading I think
There is support for that in the paid version, but not in the open version. How did you manage to achieve that?
Hi ShortElephant92 , are you referring to the open-source ClearML server?
WackyRabbit7 this is most likely a Cookie issue - your browser already had a cookie with an old Token for the previous server, and the UI failed trying to access the server using that Cookie. Clearing the cookies is always a good thing when reinstalling servers.
Hi @<1554275773496430592:profile|DeliciousRaven95> , just to make sure I understand, you wish to make the fileserver store the actual data on NFS? Is that's the case, you'll need to set a PVC for the appropriate driver (using the correct storage class)
As far as I know the automatic binding uses async upload, which should be verbose
Hi VexedCat68 , this is a reported issue with ClearML SDK 1.1.3 - you can upgrade to the latest RC - 1.1.4rc0
Hi @<1554275773496430592:profile|DeliciousRaven95> , ClearML will only show registered datasets
Also in this docker-compose I removed the external binding of the ports for mongo/redis/es
Yes, latest docker-compose was already updated with this change 🙂
JitteryCoyote63 can you share the docker-compose file so we can make sure the documentation follows it?
@<1570583227918192640:profile|FloppySwallow46> can you check the instances in the AWS dashboard? is it possible they are stuck and the autoscaler cannot communicate with them?
(also, did you include the complete log?)