These are chat archives for atomix/atomix

3rd
Feb 2017
Jon Hall
@jhall11
Feb 03 2017 01:23
I’ve just been able to get some timeout exceptions after restarting nodes, but I was in a meeting so I haven’t looked too much into it yet
but it seems better than before
Jordan Halterman
@kuujo
Feb 03 2017 01:23
hmm
well let me know when you get a chance and can reproduce/elaborate on that
Jon Hall
@jhall11
Feb 03 2017 01:30
ahh, never mind, I accidently was using atomix 1.0 instead of the snapshot
Jordan Halterman
@kuujo
Feb 03 2017 01:31
hmm but that PR isn’t in the snapshot… hasn’t been merged yet
Jon Hall
@jhall11
Feb 03 2017 01:32
well it is locally :)
Jordan Halterman
@kuujo
Feb 03 2017 01:32
Haha gotcha
Duh
Jon Hall
@jhall11
Feb 03 2017 01:35
Someone else may have come been able to reproduce it, but without debug logs, we couldn’t tell the cause of the issue, and I think there was at least one known bug they ran into. They are rerunning with debug enabled, but I haven’t heard back from them
Jordan Halterman
@kuujo
Feb 03 2017 01:35
awesome
Jon Hall
@jhall11
Feb 03 2017 01:35
but other than that, my tests ran all last night without hitting any store issues
Jordan Halterman
@kuujo
Feb 03 2017 01:43
really need to get the Jepsen tests extended to use the more complex portions of the Copycat protocol to try to weed out these problems
and make them part of the CI
Jon Hall
@jhall11
Feb 03 2017 01:45
I played around with those briefly, but they didn’t play well in my vm
Jordan Halterman
@kuujo
Feb 03 2017 01:45
yeah I think they need a lot of work
we’ve really only run them at certain points in the development process, and not for a while
they really need to be made part of normal testing
this would be a perfect case for them