New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
journald misses initial kernel messages on arm64 #11104
Comments
Test failed in #11099, see "bionic-arm64 — autopkgtest finished (failure)"
Full log: |
Another example:
|
It could be a problem with the way the test is implemented. If this is the case, class JournalTest(unittest.TestCase):
'''Check journal functionality'''
def test_no_options(self):
out = subprocess.check_output(['journalctl'])
# has kernel messages
self.assertRegex(out, b'kernel:.*[cC]ommand line:')
# has init messages
self.assertRegex(out, b'systemd.*Reached target Graphical Interface')
# has other services
self.assertRegex(out, b'NetworkManager.*:.*starting') Specifically, kernel message regex can be loosened a bit so that it matches |
Noticed the same problem in another PR #11100. Same arch - "bionic-arm64 — autopkgtest finished (failure)".
Surprisingly to me, the tests aren't run in a clean environment: "Logs begin at" date is the same as in my previous example. |
Third data point. PR #11165. The test and the result are the same.
Again, "Logs begin at" date is the same as in other examples. |
Fourth data point. PR #11177. The same test failed.
This time, "Logs begin at" date is different, so the test is probably running |
bionic-arm64 has just been turned off entirely but journald is still missing messages so let's keep the issue open. |
To get this test case passing on arm64 I did short-circuit the test for kernel log in the Ubuntu builds. I should upstream this, but also it is worrying that journald does consistently miss early kernel kmsg on arm64. I wonder if the buffer is too small, or there are too many kmsg by the time journald starts, or something like that. |
Is this the kernel discarding log messages or journald? If the former, might |
@mbiebl I'm wondering if you have ever seen this on Debian CI on arm. Frankly, I'm inclined to forward the issue to https://bugs.launchpad.net because I don't think there is anything we can do here to even debug this without access to the Ubuntu infrastructure. |
Judging by @xnox's comment it seems to be a known issue that was "resolved" somewhere so apparently I don't even need to file an issue on https://bugs.launchpad.net/. I'm going to close it here. |
well, but it's not yet in bionic kernel, so not quite done done. But it's this one: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1824864 |
all supported series now have the log_buf_len kernel config option bumped up. |
Good to know. The test itself was fixed several months ago and |
systemd version the issue has been seen with
Expected behaviour you didn't see
Unexpected behaviour you saw
Steps to reproduce the problem
Run CI autopkgtest, case
JournalTest.test_no_options
.The text was updated successfully, but these errors were encountered: