Am 26.08.26 um 07:47 schrieb Jochen Sprickerhof:
* Michael Biebl <[email protected]> [2026-08-25 23:04]:Am 25.08.26 um 20:12 schrieb Michael Biebl:Am 25.08.26 um 20:09 schrieb Michael Biebl:Hi Jochen,is there a way to find this build failure from the tracker.d.o dashboard?It is listed right there:Reproducibility regression on amd64: rsyslog, rsyslog-clickhouse, rsyslog-czmq, rsyslog-docker, rsyslog-elasticsearch, rsyslog-gnutls, rsyslog-gssapi, rsyslog-hiredis, rsyslog-http, rsyslog-kafka, rsyslog- kubernetes, rsyslog-mongodb, rsyslog-mysql, rsyslog-openssl, rsyslog- pgsql, rsyslog-relp, rsyslog-snmp
Hm, I could swear I didn't see this in the excuses output the other day. But now it is there indeed.
Can the package be given back / retried?It was retried multiple times already and failed every time.Problem is, the test suite passes really reliably here on my local system, i.e. I can reproduce the problem or create an environment whereI can *not* reproduce the problem...I tried building it using debrebuild --builder=sbuild+unshare and debrebuild --builder=sbuildFor completeness, reproduce.debian.net uses --builder=sbuild+unshare but that is actually unrelated as I can get the same result with:TMPDIR=/var/tmp sbuild --chroot-mode=unshare --dist=unstable rsyslog But only on some computers..
I've only ever see this particular test fail on reproducible-builds.org and now on reproduce.debian.net
You said you ran your own tests, what were your findings so far?I only found that it happens when compiling on ext4 (via setting TMPDIR as above). Do you need anything specific from a failed run to debug this?
Unless I find the circumstances / environment how this can be reproduced, I think there is no real point in forwarding this issue upstream.
Tbh, I'm tempted to simply disable 'make check' when it is run as part of a reproducible build. Is there a reliable way to detect this from within debian/rules?
OpenPGP_signature.asc
Description: OpenPGP digital signature
