Learn more about these different git repos.
Other Git URLs
We've seen this from time to time where binary data gets into the logs for creating ostrees
From last night's rawhide run: - x86_64 task log - ppc64le task log
From last night's f26 run: - x86_64 task log
<img alt="rawhide-x86_64-root.log" src="/pungi/issue/raw/fabc02205b5e6ad31c9ffa700073cdbec04e0e211a668867428f6538e49fe541-rawhide-x86_64-root.log" /> <img alt="rawhide-x86_64-root.log" src="/pungi/issue/raw/fabc02205b5e6ad31c9ffa700073cdbec04e0e211a668867428f6538e49fe541-rawhide-x86_64-root.log" /> <img alt="rawhide-x86_64-root.log" src="/pungi/issue/raw/fabc02205b5e6ad31c9ffa700073cdbec04e0e211a668867428f6538e49fe541-rawhide-x86_64-root.log" />
This could be pungi or rpm-ostree but being that the binary data only starts coming to the log after rpm-ostree has completed it looks like it's not that. This could be a bad metric because perhaps it's buffering that makes it look like it only comes after rpm-ostree has completed. Basically this is the last line that is printed before the binary data starts spewing out in these 3 cases, let's start there.
I know this happens every now and then. There is #584 (now closed as duplicate of this issue with more info).
The Updating ref line is the last thing that is supposed to be printed. When the binary data appears on failed runs, it's after the traceback from pungi-make-ostree. That makes me think it's not an issue with the script itself.
Updating ref
pungi-make-ostree
The data is also present in mock_output.log.
mock_output.log
Since this only happens for ostree tasks, I'm wondering if this could be related to the new_chroot option.
new_chroot
ok so there was a build yesterday - from the directory structure this runroot.log says that the koji task 20127733 is related to this build.
From the same directory structure there is this create-ostree-repo.log that does not have any binary data in it, but from the koji task logs this root.log has binary data in it.
That would indicate to me that it's not rpm-ostree/ostree that is sending binary data to the logs, but something else. Is that safe to say?
<img alt="runroot.log" src="/pungi/issue/raw/942adbfc7c8dc5cb18c720eeaafc85d9d58d278e00a38b82e8763d3fb5725181-runroot.log" /> <img alt="root.log" src="/pungi/issue/raw/c396deda4252b67019c1e5a1e9b4ef8f374cf7eda13284096e4e32cfa905122e-root.log" /> <img alt="create-ostree-repo.log" src="/pungi/issue/raw/5244694a82fe5798a52c30c6307fe9562dcf61dfe62891b84f9bc8fd6a479c66-create-ostree-repo.log" />
Your analysis seems to be correct. The create-ostree-repo.log is where output of rpm-ostree is redirected. Again the binary data is printed after the pungi command exits.
create-ostree-repo.log
rpm-ostree
I filed koji#482. Maybe koji devs can help point us in the right direction.
I tried running re-running 20127733 in mock locally, and not only there is no binary data, it also succeded (maybe some change in the buildroot since the task ran).
unfortunately this isn't consistent. It only happens randomly which is why you see me pop back into this issue periodically. I'm just trying to give as much information as possible so we can run down the issue eventually.
@dustymabe Is this still happening?
I don't think there's anything we can do about this now, and as far as I know it stopped happening. I'm going to close this issue.
Metadata Update from @lsedlar: - Issue close_status updated to: Insufficient data - Issue status updated to: Closed (was: Open)
+1 - we haven't seen this for a while.
Login to comment on this ticket.