Your message dated Sat, 21 Mar 2026 18:18:59 +0000
with message-id <[email protected]>
and subject line Bug#112696: fixed in nginx 1.28.2-3
has caused the Debian Bug report #112696,
regarding doesn't write out ta while lck remains set - data loss happens
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
112696: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=112696
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Package: net-acct
Version: 0.7-2
Severity: important
Hi,
we are using net-acct for our ISP accounting system to collect the
data. We have been using net-acct for quite some time on a variety of
systems. Recently, we have been experiencing problems that net-acct
stops writing out data and takes all available CPU.
With debugging options set, we get logs that look like this:
18/09 21:06:24 [MISC ] Total of 24464 entries
18/09 21:06:24 [STATE] dump child: synchronized with parent
18/09 21:06:24 [STATE] * dump process 10525 forked
18/09 21:06:24 [STATE] parent: synchronized with dump child
18/09 21:06:24 [STATE] dumppid is 10525
18/09 21:06:24 [STATE] lck = 0
18/09 21:06:24 [MISC ] Split into 8849 [hold] and 15615 [write] = 24464 [total]
entries
18/09 21:06:24 [STATE] write child: synchronized with parent
18/09 21:06:24 [STATE] * write process 10526 forked
18/09 21:06:24 [STATE] parent: synchronized with write child
18/09 21:06:24 [STATE] writepid is 10526
18/09 21:06:24 [STATE] done freeing
18/09 21:06:25 [STATE] * dump finished, count = 8849
18/09 21:06:25 [STATE] -> got signal 17, handling
18/09 21:06:25 [SIG ] waitpid returned 10525, status = 0, errno = 4
18/09 21:06:25 [STATE] set dumping to 0
18/09 21:06:25 [STATE] child 10525 signaled return
18/09 21:06:25 [STATE] <- got signal 17, done handling
18/09 21:06:25 [STATE] * write finished, count = 15615
18/09 21:06:25 [STATE] -> got signal 17, handling
18/09 21:06:25 [SIG ] waitpid returned 10526, status = 0, errno = 4
18/09 21:06:25 [STATE] set writing to 0
18/09 21:06:25 [STATE] child 10526 signaled return
18/09 21:06:25 [SIG ] waitpid returned -1, status = 0, errno = 10
18/09 21:06:25 [STATE] <- got signal 17, done handling
18/09 21:07:24 [STATE] write_log called
18/09 21:07:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:08:25 [STATE] write_log called
18/09 21:08:25 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:09:24 [STATE] write_log called
18/09 21:09:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:10:24 [STATE] write_log called
18/09 21:10:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:11:24 [STATE] write_log called
18/09 21:11:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:12:24 [STATE] write_log called
18/09 21:12:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:13:24 [STATE] write_log called
18/09 21:13:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:14:24 [STATE] write_log called
18/09 21:14:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:15:24 [STATE] write_log called
18/09 21:15:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:16:24 [STATE] write_log called
18/09 21:16:24 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:17:25 [STATE] write_log called
18/09 21:17:25 [STATE] flushing delayed (writing == 0, lck == 1, may_write == 1)
18/09 21:18:24 [STATS] ignored: 2889478 netignored: 0 local:0 ip:1340990
unenc:0 notdev:0 dropped:0
18/09 21:18:24 [STATS] udp: 236757 tcp:1075249 icmp:24883 other:4101
This usually repeats ad infinitum, but in this rare case, the daemon
got back on track, eventually writing out the data. We have inspected
the code and didn't find any occurences where lck is written to
without debug output, but obviously, lck was set to 1 somewhen between
21:06:24 and 21:07:24, and reset between 21:17:25 and 21:18:24.
This behavior can be reproduced by setting "large" timeout values for
flush (300) and fdelay (120).
We successfully backported sid's net-acct 0.71-2 to potato, but the
bug shows here as well.
There does not seem to be a public mailing list about net-acct, so the
BTS probably is the only way to get in touch with knowledgeable
people. Any ideas what might be going wrong here?
Greetings
Marc
-- System Information
Debian Release: 2.2
Architecture: i386
Kernel: Linux paola 2.2.19 #1 Wed Apr 18 19:00:04 CEST 2001 i686
Versions of packages net-acct depends on:
ii cron 3.0pl1-57.3 management of regular background p
ii libc6 2.1.3-18 GNU C Library: Shared libraries an
-- Configuration Files:
/etc/naccttab changed [not included]
--- End Message ---
--- Begin Message ---
Source: nginx
Source-Version: 1.28.2-3
Done: Jan Mojžíš <[email protected]>
We believe that the bug you reported is fixed in the latest version of
nginx, which is due to be installed in the Debian FTP archive.
A summary of the changes between this version and the previous one is
attached.
Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to [email protected],
and the maintainer will reopen the bug report if appropriate.
Debian distribution maintenance software
pp.
Jan Mojžíš <[email protected]> (supplier of updated nginx package)
(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing [email protected])
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
Format: 1.8
Date: Sat, 21 Mar 2026 17:37:50 +0000
Source: nginx
Architecture: source
Version: 1.28.2-3
Distribution: unstable
Urgency: medium
Maintainer: Debian Nginx Maintainers
<[email protected]>
Changed-By: Jan Mojžíš <[email protected]>
Closes: 112696
Changes:
nginx (1.28.2-3) unstable; urgency=medium
.
* All changes in this release focus on mitigating incorrect Host header
handling in proxied requests by backporting $request_port and
$is_request_port variables from newer nginx, and explicitly setting
HTTP_HOST in proxy_params, fastcgi_params, uwsgi_params and
scgi_params to $host$is_request_port$request_port. (Closes: 112696)
* d/p/debian/patches/backport-request_port_and_is_request_port.patch: add
- backport of $request_port and $is_request_port variables from newer nginx
- implements support across HTTP/1.x as well as HTTP/2 and HTTP/3 code
paths
- note: HTTP/3 module in nginx is still considered experimental upstream
* d/conf/proxy_params: use $host$is_request_port$request_port instead of
$http_host
* d/t/proxy: add test verifying $host$is_request_port$request_port behavior
* d/t/fastcgi-RFC9112: add test that verifies the correct Host value
according to RFC 9112, thanks to @yadd
* d/conf/fastcgi_params:
add 'fastcgi_param HTTP_HOST $host$is_request_port$request_port;'
* d/t/uwsgi-RFC9112: add test that verifies the correct Host value
according to RFC 9112 for uwsgi
* d/conf/uwsgi_params: add
'uwsgi_param HTTP_HOST $host$is_request_port$request_port;'
* d/conf/scgi_params: add
'scgi_param HTTP_HOST $host$is_request_port$request_port;'
Checksums-Sha1:
18edccc1c6f4b097bfeff81f986d9d5451443ef3 3803 nginx_1.28.2-3.dsc
afc4e0aecf54fccfa4150ae17b022c84f2d3ace2 75880 nginx_1.28.2-3.debian.tar.xz
22bb9c15a9f210e52bc10e0ee584e784208ab2a4 3548320 nginx_1.28.2-3.git.tar.xz
70488a999ef28ff14fc2eb614d8696b47de1eaea 17327 nginx_1.28.2-3_source.buildinfo
Checksums-Sha256:
f4d4608ede8b700231cc6da18d2b6d6983fff45c0d86be9babfa06c98d9dabf0 3803
nginx_1.28.2-3.dsc
eb04ad525fbf08e02fb51f09b5428e76a593b10800fa8a369431673bb1ec6e33 75880
nginx_1.28.2-3.debian.tar.xz
08a797694d23adfafb994a47700e2d748cc66782f1b13d18b752e138067c4f81 3548320
nginx_1.28.2-3.git.tar.xz
17abe2c241b61955175db446577acba76d6ae96b175ec0d48b3566a9bab0aa1b 17327
nginx_1.28.2-3_source.buildinfo
Files:
0b0c6d6baad389f46a78fbb8d14bfbd9 3803 httpd optional nginx_1.28.2-3.dsc
6beb7b44cfa05a89050ecb53e292de48 75880 httpd optional
nginx_1.28.2-3.debian.tar.xz
a582d970be47ea1d3e63e7cc1375ce0d 3548320 httpd None nginx_1.28.2-3.git.tar.xz
0c1c9616e5b7938f19eece20343a9127 17327 httpd optional
nginx_1.28.2-3_source.buildinfo
Git-Tag-Info: tag=584014647d1207283b2224f2ab2550fdbbe1d3e9
fp=d008b0c23d8479e46b9fcb9045da517496939ff9
Git-Tag-Tagger: Jan Mojžíš <[email protected]>
-----BEGIN PGP SIGNATURE-----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=qGlL
-----END PGP SIGNATURE-----
pgpvOzjTGIsGE.pgp
Description: PGP signature
--- End Message ---
