Jobs are failing with the following error in n-cpu:
Guest refused to detach volume <uuid>:
nova.exception.DeviceDetachFailed: Device detach failed for vdb:
Unable to detach the device from the live config.
At this time, this query has:
20 hits in the last 7 days, check and gate, all failures
Related-Bug: #1882521
Change-Id: Ib92b679f2d1dbd8131f58c8bb85fc2a3f65dbfb5
The previous query for 1708704 was causing e-r to OOM and resulting in a
lack of graph data. This is beacuse tripleo's logstash.txt file is not
getting parsed properly and ends up with the entire file as the single
event. Then e-r downloads all those file copies and fills its memory and
breaks. We can work around this by only looking in job-output.txt files
for this bug. Then when tripleo's fix has flushed the bad events out (10
days after fix merges) we can revert this change.
Change-Id: Id619f90ffe84b3d4de334ea4b17026b9b3239d33
45 hits in the last 7 days, with a spike since Dec 16, mostly
fortnebula nodes, all failures, check and gate.
Change-Id: Ic856d11e183075244de556b86d4ecdc7bcc78abd
This also shows up in multi-node jobs when attaching a volume
so update the query to include volume attach failures.
Change-Id: Ie77e3998b2ff4a508fa3b6078acb3ceec15d7e37
211 hits in 7 days, check and gate, all failures.
The message shows up in the n-api logs but filtering
on n-api logs isn't sufficient to get 100% failure rate
in logstash because there are tempest tests like
test_create_list_show_delete_interfaces_by_network_port
which handle the error and work around it so even though
the message shows up not all jobs fail because of it so
we use the tempest failure log in the console to fingerprint
this bug.
Change-Id: I7b4a3f4a483c5166e9aee1507f12bb31069a8fe0
This is split off from bug 1848078 since this test specifically
seems to hit this issue a lot in multinode jobs where the instance
is shelved from one node and unshelved on another.
14 hits in 7 days, all failures.
Change-Id: I9bd41e356abf72ff08415693b7b2b11a035a542d
26 hits in 7 days, check and gate, all failures. Seems
to be restricted to stable/rocky for some reason.
Change-Id: Ic9463841764c5c4dc552e0b2332d854234b609f5
1278 hits in 24 hours since networkx 2.4 was released.
In nova's case the problem is upper-constraints weren't
being applied to runtime dependencies or transitive
dependencies in tox.ini.
Change-Id: Ie15f3ebbe33116704fa6e066933e69bb6b170f33
6 hits in 7 days, check queue only right now but multiple
changes, all failures since it breaks SSHing into the guest
after it's hard-rebooted.
Change-Id: Ic9be31f67af3796760b0d514a903f5e9bcd4aca1
It has a low hit rate because it's a race in a very specific case
(deleting an instance before it finished building), but it's a
legitimate bug nonetheless.
Change-Id: Ib5014af4518f7682b79fdc85b7af06814747bacf
We hit this failure sometimes trying to get other metadata
API routes besides meta-data, e.g. user-data:
failed to get http://169.254.169.254/2009-04-04/user-data
This makes the query more generic to get more hits. This
brings the hit count up to 86 in voting jobs in the last
10 days, multiple branches, check and gate, all failures.
Looks like "ovh-bhs1" is the top offender node provider.
Change-Id: I9fb3885fdc674c7176fc2439a2bbdb65b84048a1
1101 hits since Sept 17, check and gate, all failures. Those
numbers are a bit inflated since the message gets duplicated
while the problem occurs.
Change-Id: I12e174effad802928a01800df975990acc908c7a
This is pretty rare, only 1 hit in 7 days but it's in the gate queue
and is a known issue (the bug was reported against stein) so we might
as well track it.
Change-Id: Ic36fa2b1afe11bdea7af30b897dcc242e442582f
357 hits in 48 hours, check and gate, all failures. It looks
like this is resolved now so this is just getting it off the
uncategorized bugs page.
Change-Id: I88a7137c7dfa1fb50a2385d6a8fa74565348dbc3
Related-Bug: #1843715
Some of these are removed by the cleanup script
and some were removed by hand since they were
duplicates of invalid or fixed bugs and the cleanup
script doesn't yet handle chasing down duplicates.
Change-Id: I2f691511fe620654d5949d617f8131c1360949f7
This is pretty rare in the nova-grenade-live-migration job so right now
it only has 1 hit in 10 days in the check queue, but I've seen it before
so wanted to start tracking it.
Change-Id: I11011f2d60455721490c48fae17c39ffafe28264
This is an old bug that still shows up in multi-node
jobs. 60 hits in the last 7 days (but logstash seems
to be a bit off so there are probably more), all failures.
Change-Id: Ic6135ddef4e503fa1618d5a4fe4fd04309f18939
44 hits since https://review.opendev.org/#/c/203698/ merged,
check and gate, all failures, but only in neutron-functional
and fullstack type jobs, I'm not sure why.
Change-Id: Ic2ef0281481f33f10b2b563bd15fa59c191ab1ea