Hello SWAN team,
I’m trying to start a Dask cluster on the HTCondor pool using the “Bleeding Edge” software stack, and it’ won’t start: when I click on “new” I get this “Cluster Start Error” that says “Unhandled error”.
I need LCG_108, but it looks like the latest working stack with Dask is LCG_107.
Cheers,
Silia
Dear Silia,
Indeed Bleeding edge is broken due to a bug introduced by a Dask update in that stack.
We are already testing a fix that we will deploy in production soon. Please follow this issue.
Cheers,
Enric
Dear Enric,
I saw that the issue you referenced is marked as done, however, the problem persists. 
Is this still tracked somewhere else?
Cheers,
Leonhard
Dear Leonhard,
The issue referenced by Enric before was to test the changes to the dask extension in QA. We will proceed with deploying such changes in production (and the new version of the extension as well) next monday (29th of September)! You can track the progress here: https://its.cern.ch/jira/browse/SWAN-456
Best,
Pedro
