lahirujayathilake opened a new issue, #555:
URL: https://github.com/apache/airavata-custos/issues/555

   ## Problem
   
   When a user's allocation doesn't appear (e.g. cluster tooling shows "No 
Allocation", or the portal shows a membership but the cluster side is dead), 
there is no user- or admin-facing signal explaining *why*. Diagnosis today 
requires SSH-ing to the host, grepping the SLURM association-mapper log, 
correlating a per-membership "connection refused" with one specific user, and 
checking the DB by hand.
   
   Concrete case that exposed this: the cluster's `slurmrestd` was unreachable, 
so the SLURM associations were never written. The affected users could SSH in 
(their POSIX account had provisioned fine) but had no allocation on the cluster 
— with zero surfaced explanation.
   
   A user's account-to-allocation connection has two independent provisioning 
dimensions (POSIX/SSH via the identity provisioner, and the SLURM association 
via the association-mapper) that fail separately, and neither surfaces its 
state anywhere a user or admin can see.


-- 
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