Artsiom Mishuta comments

Results 103 comments of


                                            Artsiom Mishuta

scylla stuck/coredump during node start with new number of shards, after cluster was stopped

list of tests that failed by this issue not in Next_gating, I put it in queue as P3 priority

[Raft Topology] 'OperationTimedOut' error in SELECT query during adding new node

@kbr-scylla 1rst run - https://jenkins.scylladb.com/view/staging/job/scylla-staging/job/artsiom_mishuta/job/dtest-pytest-gating/255/ 14 fails for 200 runs with error cassandra.OperationTimedOut 2nd run https://jenkins.scylladb.com/view/staging/job/scylla-staging/job/artsiom_mishuta/job/dtest-pytest-gating/256/ - 15 failures

[Raft Topology] 'OperationTimedOut' error in SELECT query during adding new node

@kbr-scylla tested on scylla-unified-5.5.0~dev-0.20240319.a6441b7fe9b3 - 0 failures

[Raft Topology] 'OperationTimedOut' error in SELECT query during adding new node

update PR to enable test fo raft topology https://github.com/scylladb/scylla-dtest/pull/3964

Write handler blocks topology barriers for the duration of the timeout if live host is detected by gossip as having restarted

@tgrabiec, is this issue still actual or fix https://github.com/scylladb/scylladb/pull/17585 solved the problem?

test.py: teardown may race with ongoing test operation, causing IPs to be released prematurely

@gleb-cloudius, @kbr-scylla, Do you have a reproducer for this, some test example or an existing test? I haven't worked with test.py but to deal with concurrent threads/await in Python you...

test.py: teardown may race with ongoing test operation, causing IPs to be released prematurely

@kostja, I read issue/code one more time. and I see that ```cmd``` it is already sync point between ```start``` ```stop``` and cmd filed as ``` self.cmd = await asyncio.create_subprocess_exec``` almost...

test.py: teardown may race with ongoing test operation, causing IPs to be released prematurely

@xtrey, I've suggested "synchronization on server" because the initial issues mentioned race between 2 meant that performing an API call on server (start, remove) but @kbr-scylla example as only coro...

fix(upgrade): upgrade with raft topology procedure

blocked by https://github.com/scylladb/scylladb/issues/18098

Alter the restore nemesis for raft (and tablets) scenarios

reproduced in https://argus.scylladb.com/test/379b8fb7-0e42-4453-bc77-b857a8758986/runs?additionalRuns[]=c2048c63-a1d8-44a2-b0d9-8874c6d9b145 2024-07-22 09:48:43