Page MenuHomePhorge

Commits in Phorge stuck on 'Still Importing'
Closed, InvalidPublic

Description

Copied from Q106

I've recently setup an instance of Phorge running in a jail (FreeBSD) of a TrueNAS installation. Everything appeared to be working for some weeks, but a few days ago new commits appear stuck on 'Still Importing'.
Visitting the /daemon page on the WebUI, for each offending commit an entry appears under 'Leased Tasks', but for each of those the Failures count remains at 0.
I've updated my install to the latest available on the master branch, but it has not changed the observed behavior.

output of /bin/repository importing:

/phorge # ./bin/repository importing rMA
rMA5ec5cf9d327b Publish
rMAc68f8ce88ce2 Publish
... (omitted a few more)
rMA4ec5bba1ba2a Publish

When trying to reparse one of the broken commits, the following output is observed:
(without --trace)

/phorge # ./bin/repository reparse --importing rMA4ec5bba1ba2a
[                                                                  ]   0.0%Abort

when running this command with --trace, the following output appears:

... (omitted as things seem to be working initially - please ignore the misspelled 'deamonuser')
>>> [32] (+103) <exec> $ which sudo
<<< [32] (+113) <exec> 9,388 us
>>> [33] (+113) <exec> $ /usr/local/bin/sudo -E -n -u deamonuser -- git log -n 1 '--format=%P' 4ec5bba1ba2a943726017956042c48b0bc24472d --
<<< [33] (+132) <exec> 18,797 us
>>> [34] (+133) <exec> $ which sudo
<<< [34] (+140) <exec> 7,077 us
>>> [35] (+161) <query> SELECT * FROM `phabricator_file`.`file` WHERE contentHash = 'fbc3a6580e6f6567970a6b6bf3c9103137453cec9351d71e2c57b3789f600a76' LIMIT 1
<<< [35] (+162) <query> 832 us
>>> [36] (+165) <kvcache-get>
>>> [37] (+165) <connect> phabricator_cache
<<< [37] (+166) <connect> 983 us
>>> [38] (+167) <query> SELECT * FROM `cache_general` WHERE cacheKeyHash IN ('BnTlrCiB63YN')
<<< [38] (+167) <query> 444 us
<<< [36] (+167) <kvcache-get> 2,157 us
>>> [39] (+167) <kvcache-get>
>>> [40] (+167) <query> SELECT * FROM `cache_general` WHERE cacheKeyHash IN ('BnTlrCiB63YN')
<<< [40] (+168) <query> 291 us
<<< [39] (+168) <kvcache-get> 466 us
>>> [41] (+168) <connect> phabricator_file
<<< [41] (+169) <connect> 845 us
>>> [42] (+169) <query> INSERT INTO `phabricator_file`.`file_storageblob` (`data`, `dateCreated`, `dateModified`) VALUES ('diff --git a/test.txt b/test.txt\nnew file mode 100644\nindex 000000000..980a0d5f1\n--- /dev/null\n+++ b/test.txt\n@@ -0,0 +1 @@\n+Hello World!\n', '1706183654', '1706183654')
<<< [42] (+482) <query> 312,891 us
Abort

No error appears in the error logs of MySQL, and the insert query does appear in the general log if I turn on this feature.

As I am not certain what triggered this behavior in the first place, I am not sure if I could supply valid reproduction steps. That is why I would greatly appreciate any advice on what could be causing this, or suggestions on how to figure out the root cause of this problem.

Thanks in advance!

Event Timeline

avivey subscribed.

Hi,
this should probably be a question in Ponder, not a task/bug report.
Bug reports require reproduction.

This also probably doesn't effect FreeBSD the organization.