Mailman-web stop responding.
I was trying to administrate a list and discovered that mailman-web was unresponsive.
I have tried stopping and restarting it. The error logs update for the startup, but not when anything connects to localhost:8000. The last entry in uwsgi.log was two hours ago and nothing has been added since.
I tried both lynx http://localhost:8000 and telnet to http://localhost 8000 and GET request instantly closes the port and logs nothing.
As far as I can tell both mailman and postgres are working fine. This is what logs on startup.
==> uwsgi-error.log <== *** Starting uWSGI 2.0.22 (64bit) on [Sun Apr 14 10:45:51 2024] *** compiled with version: 13.2.1 20230912 [revision b96e66fd4ef3e36983969fb8cdd1956f551a074b] on 05 October 2023 13:37:28 os: Linux-6.5.4-1-default #1 SMP PREEMPT_DYNAMIC Wed Sep 20 05:07:04 UTC 2023 (fdd7e9e) nodename: mail.rubinsoftware.com machine: x86_64 clock source: unix detected number of CPU cores: 4 current working directory: / detected binary path: /opt/mailman/venv/bin/uwsgi !!! no internal routing support, rebuild with pcre support !!! your processes number limit is 15246 your memory page size is 4096 bytes detected max file descriptor number: 1024 lock engine: pthread robust mutexes thunder lock: disabled (you can enable it with --thunder-lock) uwsgi socket 0 bound to TCP address 0.0.0.0:8000 fd 8 Python version: 3.11.5 (main, Aug 31 2023, 07:57:41) [GCC] PEP 405 virtualenv detected: /opt/mailman/venv/ Set PythonHome to /opt/mailman/venv/ Python main interpreter initialized at 0x7f2c3d709d58 python threads support enabled your server socket listen backlog is limited to 100 connections your mercy for graceful operations on workers is 60 seconds mapped 250128 bytes (244 KB) for 4 cores *** Operational MODE: preforking+threaded *** added /etc/mailman3/ to pythonpath. WSGI app 0 (mountpoint='') ready in 0 seconds on interpreter 0x7f2c3d709d58 pid: 16064 (default app) *** uWSGI is running in multiple interpreter mode *** spawned uWSGI master process (pid: 16064) spawned uWSGI worker 1 (pid: 16065, cores: 2) spawned uWSGI worker 2 (pid: 16067, cores: 2) [uwsgi-daemons] spawning "/opt/mailman/venv/bin/mailman-web qcluster" (uid: 72 gid: 67) 14:45:52 [Q] INFO Q Cluster white-green-ten-mobile starting. 14:45:52 [Q] INFO Process-1:1 ready for work at 16072 14:45:52 [Q] INFO Process-1:2 ready for work at 16073 14:45:52 [Q] INFO Process-1:3 monitoring at 16074 14:45:52 [Q] INFO Process-1 guarding cluster white-green-ten-mobile 14:45:52 [Q] INFO Process-1:4 pushing tasks at 16075 14:45:52 [Q] INFO Q Cluster white-green-ten-mobile running.
What should I do to diagnose further?
I just thought to look at top. Mailman-web is the top process and bouncing between 30 and 1% it only gets pushed off the top when new mail is received and immediately returns.
Paul Rubin wrote on 2024-04-14 08:06:
detected binary path: /opt/mailman/venv/bin/uwsgi !!! no internal routing support, rebuild with pcre support !!!
I don't know about your problem with it not being responsive, but this message can be solved by running (inside the venv, assuming Ubuntu / Debian OS):
pip uninstall uwsgi sudo apt install libpcre3 libpcre3-dev pip install --force-reinstall -U --no-deps --no-cache-dir uwsgi
No idea how important PCREs are to Mailman3, but it probably doesn't hurt to have it.
It isn't mentioned in the venv install guide.
The package library does not have that lib. When I reinstalled uwsgi, it when from 2.0.22 to 2.0.25
New start log:
*** Starting uWSGI 2.0.25 (64bit) on [Sun Apr 14 15:40:10 2024] *** compiled with version: 13.2.1 20230912 [revision b96e66fd4ef3e36983969fb8cdd1956f551a074b] on 14 April 2024 19:38:39 os: Linux-6.5.4-1-default #1 SMP PREEMPT_DYNAMIC Wed Sep 20 05:07:04 UTC 2023 (fdd7e9e) nodename: mail.rubinsoftware.com machine: x86_64 clock source: unix pcre jit disabled detected number of CPU cores: 4 current working directory: / detected binary path: /opt/mailman/venv/bin/uwsgi your processes number limit is 15246 your memory page size is 4096 bytes detected max file descriptor number: 1024 lock engine: pthread robust mutexes thunder lock: disabled (you can enable it with --thunder-lock) uwsgi socket 0 bound to TCP address 0.0.0.0:8000 fd 8 Python version: 3.11.5 (main, Aug 31 2023, 07:57:41) [GCC] PEP 405 virtualenv detected: /opt/mailman/venv/ Set PythonHome to /opt/mailman/venv/ Python main interpreter initialized at 0x7f1697b09d58 python threads support enabled your server socket listen backlog is limited to 100 connections your mercy for graceful operations on workers is 60 seconds mapped 250128 bytes (244 KB) for 4 cores *** Operational MODE: preforking+threaded *** added /etc/mailman3/ to pythonpath. WSGI app 0 (mountpoint='') ready in 0 seconds on interpreter 0x7f1697b09d58 pid: 4870 (default app) *** uWSGI is running in multiple interpreter mode *** spawned uWSGI master process (pid: 4870) spawned uWSGI worker 1 (pid: 4871, cores: 2) spawned uWSGI worker 2 (pid: 4873, cores: 2) [uwsgi-daemons] spawning "/opt/mailman/venv/bin/mailman-web qcluster" (uid: 72 gid: 67) 19:40:11 [Q] INFO Q Cluster low-helium-finch-queen starting. 19:40:11 [Q] INFO Process-1:1 ready for work at 4878 19:40:11 [Q] INFO Process-1:2 ready for work at 4879 19:40:11 [Q] INFO Process-1:3 monitoring at 4880 19:40:11 [Q] INFO Process-1 guarding cluster low-helium-finch-queen 19:40:11 [Q] INFO Process-1:4 pushing tasks at 4881 19:40:11 [Q] INFO Q Cluster low-helium-finch-queen running. invalid request block size: 21573 (max 4096)...skip
On Wed, Apr 17, 2024 at 3:43 PM <paulrubin37@gmail.com> wrote:
I have no idea what changed, but it suddenly started working.
This is not good feedback when you have all the log files to discern the issue from - unless you disabled all logging. Anyway, maybe God solved it :-)
-- Best regards, Odhiambo WASHINGTON, Nairobi,KE +254 7 3200 0004/+254 7 2274 3223 In an Internet failure case, the #1 suspect is a constant: DNS. "Oh, the cruft.", egrep -v '^$|^.*#' ¯\_(ツ)_/¯ :-) [How to ask smart questions: http://www.catb.org/~esr/faqs/smart-questions.html]
participants (4)
-
Odhiambo Washington -
Paul Rubin -
paulrubin37＠gmail.com
-
Ron / BCLUG