-
Notifications
You must be signed in to change notification settings - Fork 33
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
Image refresh for arch (for fsinfo) #5861
Conversation
image-refresh arch done: https://github.com/cockpit-project/bots/commits/image-refresh-arch-20240202-202818 |
6c588a3
to
bb54e61
Compare
How very convenient. :) |
Great. More
and indeed: sendto(5, "HEAD / HTTP/1.1\r\nHost: 172.27.0.15:9090\r\nUser-Agent: curl/8.6.0\r\nAccept: */*\r\n\r\n", 80, MSG_NOSIGNAL, NULL, 0) = 80
recvfrom(5, "HTTP/1.1 301 Moved Permanently\r\nContent-Type: text/html\r\nLocation: https://172.27.0.15:9090/\r\n\r\n<html><head><title>Moved</title></head><body>Please use TLS</body></html>\r\n", 102400, 0, NULL, NULL) = 171 That looks distinctly like a body sent in reply to a HEAD request. This one would have been coming from |
The two cockpit PRs landed, retrying. I didn't yet look into the two machines failures, but we've accrued quite a number of flakes there, so retrying for comparison as well. |
Ah, there is still the testRaidRepair crash, which already fixed in the other pending refresh in PR #5804 . I suppose this is for @mvollmer. |
The same regression now landed in Fedora updates-testing, see cockpit-project/cockpit#19937 . Smells like a kernel regression? |
I wouldn't be surprised if it's related to the one in #5793... |
If you have a link to the kernel patch which is supposed to resolve it, I can easily build a kernel / verify |
The ubuntu one was a rather different area though (partition vs. RAID repair) |
I'll have a look as well. The first failure is interesting, the other two with a busy /dev/loop10 is just a failed cleanup. testRaidRepair does indeed look like https://bugzilla.redhat.com/show_bug.cgi?id=2256432, but MaxLayoutSizes must be something else... |
Yes, lvconvert hangs with the exact same kernel stack trace. |
This seems to happen pretty easily, also with biggish disks. We might want to raise the alarm on https://bugzilla.redhat.com/show_bug.cgi?id=2256432, which has not getting any reaction so far. I have found no way to continue when this happens. The lvconvert process can not be killed by any means known to man kind when it is in this state. I think we need to make this a destructive test with a naughty. |
Blocked on cockpit-project/cockpit#19940 |
577bd28
to
f996b0f
Compare
Looking into the machines failure. All of the noise about
is just noise. I see the same thing in a local run on main with the old image (which ends successfully). Checking with the new image (and expanding the error message): That sounds like more fallout from the new curl version — curl/curl#12844 |
c-machines itself, in particular its machine_install.py, doesn't call curl. But of course that could be caused by qemu/block-curl.so calling it the wrong way. The test fakes fedoraproject.org with test/files/mock-range-server.py , but that doesn't use curl at all (it's the server side). |
Downgrading curl (using 8.5.0 from https://archive.archlinux.org/packages/c/curl/) fixes the issue. |
Landed. |
@@ -0,0 +1,3 @@ | |||
Traceback (most recent call last): | |||
File "*/check-machines-create", line *, in testCreateUrlSource | |||
runner.createTest(TestMachinesCreate.VmDialog(self, sourceType='url', |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can we make that more specific somehow? Otherwise this would match whenever anything is broken with createTest
, not just when it is curls fault.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
E.g., there is also "CURL: Error opening file: OpenSSL SSL_read: SSL_ERROR_SYSCALL, errno 0" in the browser log, a few lines above.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
OK.
|
6f62375
to
7ad57cc
Compare
Let's get a working image with fsinfo inside of it.