John Spray
John Spray
This passed CI again, giving it another round of repeats.
1 debug build failed with https://github.com/redpanda-data/redpanda/issues/5433#event-7002364672 (first time we've seen that in a CI run, but is a known rare assertion when storage is slow, doesn't seem likely to be...
@dotnwat this should be good to go now
CI failure is an intermittent hang in test_cloud_storage_rpunit which I am ignoring until https://github.com/redpanda-data/redpanda/pull/7366 merges, because it touches a bunch of the potentially-responsible code.
The segment roll messages aren't necessarily worrying, although we did introduce the change to disable segment size jitter by default since this test was originally created. The request timeouts are...
> Is 5s a realistic timeout? Yes, it should be: the test is aiming to validate that redpanda copes gracefully with this client count -- it isn't meant to overwhelm...
Here's what's getting fixed in 22.3 (and backported to 22.2): - a real test for time queries. - segment indices will ignore non-data batches when indexing, unless it is the...
Reactor stalls on a debug build aren't necessarily surprising, but if it's happening on a particular test repeatedly that's probably a sign that there really is some pathological loop in...