Bug 16011 - wbinfo_lookuprids_cache test broken
Summary: wbinfo_lookuprids_cache test broken
Status: NEW
Alias: None
Product: Samba 4.1 and newer
Classification: Unclassified
Component: Test infrastructure (show other bugs)
Version: 4.24.0
Hardware: All All
: P5 normal (vote)
Target Milestone: ---
Assignee: Samba release manager
QA Contact: Samba QA Contact
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2026-02-24 12:01 UTC by Björn Jacke
Modified: 2026-06-10 08:23 UTC (History)
3 users (show)

See Also:


Attachments
backport fixes for the tests for the 4.24 and 4.23 branches (5.62 KB, patch)
2026-06-09 20:25 UTC, Björn Jacke
vl: review+
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Björn Jacke 2026-02-24 12:01:27 UTC
The last 50 lines of log messages:

[143(183)/836 at 4m6s] samba3.blackbox.smbclient_auth.plain.--option=clientntlmv2auth=no(nt4_dc)
[144(194)/836 at 4m8s] samba3.blackbox.smbclient_auth.plain.--option=clientusespnego=no(nt4_dc)
[145(205)/836 at 4m10s] samba3.blackbox.smbclient_auth.plain.--option=clientusespnego=no --option=clientntlmv2auth=no(nt4_dc)
[146(216)/836 at 4m12s] samba3.blackbox.smbclient_auth.plain.--option=clientntlmv2auth=no --option=clientlanmanauth=yes --max-protocol=LANMAN2(nt4_dc_smb1_done)
[147(227)/836 at 4m15s] samba3.blackbox.smbclient_auth.plain.--option=clientntlmv2auth=no --option=clientlanmanauth=yes --option=clientmaxprotocol=NT1(nt4_dc_smb1_done)
[148(238)/836 at 4m17s] samba3.blackbox.smbclient_machine_auth.plain(nt4_dc:local)
[149(241)/836 at 4m17s] samba3.blackbox.smbclient_ntlm.plain NT1(nt4_dc_smb1_done)
[150(246)/836 at 4m18s] samba3.blackbox.smbclient_ntlm.plain SMB3(nt4_dc)
[151(249)/836 at 4m27s] samba3.blackbox.smbclient_machine_auth.plain(nt4_member:local)
[152(252)/836 at 4m29s] samba3.blackbox.smbclient_ntlm.plain NT1(nt4_member)
[153(257)/836 at 4m30s] samba3.blackbox.smbclient_ntlm.plain SMB3(nt4_member)
[154(260)/836 at 4m30s] samba3.blackbox.smbclient_auth.plain.--option=clientntlmv2auth=no(nt4_member)
[155(271)/836 at 4m32s] samba3.blackbox.smbclient_auth.plain.--option=clientntlmv2auth=no.member_creds(nt4_member)
[156(282)/836 at 4m34s] samba3.blackbox.smbclient_auth.plain.--option=clientusespnego=no --option=clientntlmv2auth=no(nt4_member)
[157(293)/836 at 4m36s] samba3.blackbox.smbclient_auth.plain.--option=clientusespnego=no --option=clientntlmv2auth=no.member_creds(nt4_member)
[158(304)/836 at 4m39s] samba3.blackbox.smbclient_auth.plain.--option=clientusespnego=no --option=clientntlmv2auth=no -mNT1(nt4_dc_smb1_done)
[159(315)/836 at 4m41s] samba3.blackbox.smbclient_auth.plain.--option=clientusespnego=no --option=clientntlmv2auth=no -mNT1.member_creds(nt4_dc_smb1_done)
[160(326)/836 at 4m43s] samba3.blackbox.smbclient_auth.plain.(nt4_member)
[161(337)/836 at 4m45s] samba3.blackbox.smbclient_auth.plain..member_creds(nt4_member)
[162(348)/836 at 4m47s] samba3.blackbox.smbclient_auth.empty_domain.domain_creds(nt4_member)
[163(359)/836 at 4m49s] samba3.blackbox.smbclient_auth.empty_domain.member_creds(nt4_member)
[164(370)/836 at 4m51s] samba3.blackbox.smbclient_auth.dot_domain.domain_creds(nt4_member)
[165(381)/836 at 4m53s] samba3.blackbox.smbclient_auth.dot_domain.member_creds(nt4_member)
[166(392)/836 at 4m55s] samba3.blackbox.smbclient_auth.upn.domain_creds(nt4_member)
[167(403)/836 at 4m57s] samba3.blackbox.smbclient_auth.upn.member_creds(nt4_member)
[168(414)/836 at 4m59s] samba3.blackbox.printing_var_exp(nt4_dc)
[169(416)/836 at 5m4s] samba3.blackbox.smbpasswd(nt4_dc:local)
[170(419)/836 at 5m6s] samba3.blackbox.smbclient_auth.plain.ipv6(nt4_dc)
[171(432)/836 at 5m8s] samba3.blackbox.net_cred_change(nt4_member:local)
[172(436)/836 at 5m9s] samba3.wbinfo_sids_to_xids(nt4_member:local)
[173(437)/836 at 5m9s] samba.wbinfo_lookuprids_cache(nt4_member:local)
UNEXPECTED(failure): samba.wbinfo_lookuprids_cache.delete key(nt4_member:local)
REASON: Exception: Exception: delete failed

FAILED (1 failures, 0 errors and 0 unexpected successes in 0 testsuites)

There might be more detail in /m/bjacke/a424/b532176/samba-no-opath1/./bin/ab/subunit or /m/bjacke/a424/b532176/samba-no-opath1/./bin/ab/summary.
TOP 10 slowest tests
samba3.smbtorture_s3.crypt_client.OPLOCK1(nt4_dc_smb1) -> 36
samba3.smbtorture_s3.crypt_client.LOCK1(nt4_dc_smb1) -> 21
samba3.smb2.compound_find vfs_aio_pthread_async_dosmode_default2(simpleserver) -> 20
samba3.smb2.compound_find vfs_aio_pthread_async_dosmode_default1(simpleserver) -> 17
samba3.smbtorture_s3.crypt_client.OPLOCK2(nt4_dc_smb1) -> 10
samba3.smbtorture_s3.crypt_client.LOCK12(nt4_dc_smb1) -> 10
samba3.smbtorture_s3.crypt_client.IOCTL(nt4_dc_smb1) -> 8
samba3.smbtorture_s3.crypt_client.LOCK3(nt4_dc_smb1) -> 6
samba3.smbtorture_s3.crypt_client.LOCK9B(nt4_dc_smb1) -> 6
samba3.smbtorture_s3.crypt_client.LOCK9A(nt4_dc_smb1) -> 5
ERROR: test failed with exit code 1
make: Leaving directory '/m/bjacke/a424/b532176/samba-no-opath-build'

the above written path to debug this doesn't exist after the build. /m/bjacke/a424 is an empty directory.
Comment 1 Björn Jacke 2026-03-26 13:11:02 UTC
We're running into this a couple of times and just again in a private autobuild today again.

Such flaky tests delay release processes and needlessly eat ressources. Please fix this test or if you can't fix it, please add it to the list of flapping tests.
Comment 2 Björn Jacke 2026-04-09 19:43:37 UTC
Günther, Andreas, it looks like this test is mainly yours, right? Can you please have a look at the failures? This one is the test that lets autobuild fail most often.
Comment 3 Andreas Schneider 2026-04-10 13:04:42 UTC
You are aware the the test might not be the problem but other tests. Often people do not clean up probably and then the next test stumbled over it.

Also logs would be good if you can reproduce it.
Comment 4 Andreas Schneider 2026-04-10 13:49:40 UTC
I've created https://gitlab.com/samba-team/samba/-/merge_requests/4488


The last time this test failed it was a crashing smbd!
Comment 5 Björn Jacke 2026-05-12 12:10:40 UTC
any news on this? Still hitting this many times in the release branches...
Comment 6 Guenther Deschner 2026-05-12 13:37:13 UTC
We pushed a fix Andreas made to add more debugging around this, do you have an example log for us to look?
Comment 7 Volker Lendecke 2026-05-12 13:44:39 UTC
I don't have it at, but it fails in the delete test where it seems the lowlevel tdb_delete call fails. At least we get an exception with "delete failed" as a message.
Comment 8 Björn Jacke 2026-05-12 13:47:00 UTC
I mainly experience the pain in the release branches, but those debugging changes are only in master, not the in release branches...
Comment 9 Björn Jacke 2026-06-09 15:52:33 UTC
looks like 27f57d26f950c6fbe62f9b8ef8bd5fb503f587e5 is probably the fix for this, right? Thank you Volker, great catch!
Comment 10 Volker Lendecke 2026-06-09 16:05:26 UTC
It looks highly likely to be the fix, yes. It's the "delete failed" in the particular case that this is supposed to fix. Apologies for not looking up bugzilla for the Bug: line.
Comment 11 Björn Jacke 2026-06-09 20:25:34 UTC
Created attachment 19014 [details]
backport fixes for the tests for the 4.24 and 4.23 branches