We have started the migration from htcondor 7.8.5 to 8.2. But we are facing a strange error.
Some job's submissions do not start. The only error message we got is:
Reason for last match failure: submitter limit exceeded
For the migration, we configured new master, and put some computers in its pool.
So, Two pools are running at the same time: 7.8 the production pool and the 8.2 the testing one.
We defines rules about resources (only in the 7.8 pool).
We removed these rules in the 8.2 pool.
I'm wondering if there limits can not be widespread by the 7.8 to 8.2 through some king of network commucation (udp?).
Some jobs are working well but there are not using well-used resources (resources = nfs drives).
If you have any clues or leads to the problem. We will be glad to read it.
Thanks for your time.