Hi, when I go to restart uWSGI (via ‘touch ~/web/indico.wsgi’) it causes my entire indico installation to not respond (via HTTP) for a period of 5 minutes (pretty much on the dot). I don’t think this always happened but it has been going on for a while now and I’m finally trying to figure out if this is expected behavior these days or not.
Check systemctl status indico-uwsgi.service and/or journalctl -f -u indico-uwsgi.service (both as root).
You can also systemctl restart indico-uwsgi.service (as root) instead of the touch reload, but if for some reason it takes ages to start with the touch, I would expect the same problem when doing a full restart…
hmm, the systemctl restart indico-uwsgi.service (as root) was fast as expected. I don’t really see anything interesting in the journalctl when I did the touch method:
Apr 22 13:23:25 something.edu indico-uwsgi[3612413]: Wed Apr 22 13:23:25 2026 - uWSGI worker 2 error: the master disconnected from this worker. Shutting down the worker.
Apr 22 13:23:25 something.edu indico-uwsgi[3612413]: Wed Apr 22 13:23:25 2026 - mem-collector thread started for worker 2
Apr 22 13:23:26 something.edu indico-uwsgi[985]: Wed Apr 22 13:23:26 2026 - *** /opt/indico/web/indico.wsgi has been touched… grace them all !!! ***
Apr 22 13:23:26 something.edu indico-uwsgi[985]: …gracefully killing workers…
Apr 22 13:23:26 something.edu indico-uwsgi[3557490]: Gracefully killing worker 3 (pid: 3557490)…
Apr 22 13:23:26 something.edu indico-uwsgi[3557487]: Gracefully killing worker 1 (pid: 3557487)…
Apr 22 13:23:26 something.edu indico-uwsgi[3557491]: Gracefully killing worker 4 (pid: 3557491)…
Apr 22 13:23:26 something.edu indico-uwsgi[3612413]: Gracefully killing worker 2 (pid: 3612413)…
Apr 22 13:23:27 something.edu indico-uwsgi[985]: worker 2 buried after 1 seconds
Apr 22 13:23:28 something.edu indico-uwsgi[985]: worker 1 buried after 2 seconds
Apr 22 13:23:28 something.edu indico-uwsgi[985]: worker 3 buried after 2 seconds
Apr 22 13:23:28 something.edu indico-uwsgi[985]: worker 4 buried after 2 seconds
Apr 22 13:23:28 something.edu indico-uwsgi[985]: binary reloading uWSGI…
Apr 22 13:23:28 something.edu indico-uwsgi[985]: chdir() to /
Apr 22 13:23:28 something.edu indico-uwsgi[985]: closing all non-uwsgi socket fds > 2 (max_fd = 1024)…
Apr 22 13:23:28 something.edu indico-uwsgi[985]: found fd 6 mapped to socket 0 (/opt/indico/web/uwsgi.sock)
Apr 22 13:23:28 something.edu indico-uwsgi[985]: running /opt/indico/.venv/bin/uwsgi
Apr 22 13:23:28 something.edu indico-uwsgi[985]: [uWSGI] getting INI configuration from /etc/uwsgi-indico.ini
Apr 22 13:23:28 something.edu indico-uwsgi[985]: *** Starting uWSGI 2.0.31 (64bit) on [Wed Apr 22 13:23:28 2026] ***
Apr 22 13:23:28 something.edu indico-uwsgi[985]: compiled with version: 11.4.0 on 18 December 2025 18:25:48
Apr 22 13:23:28 something.edu indico-uwsgi[985]: os: Linux-5.15.0-173-generic #183-Ubuntu SMP Fri Mar 6 13:29:34 UTC 2026
Apr 22 13:23:28 something.edu indico-uwsgi[985]: nodename: something.edu
Apr 22 13:23:28 something.edu indico-uwsgi[985]: machine: x86_64
Apr 22 13:23:28 something.edu indico-uwsgi[985]: clock source: unix
Apr 22 13:23:28 something.edu indico-uwsgi[985]: pcre jit disabled
Apr 22 13:23:28 something.edu indico-uwsgi[985]: detected number of CPU cores: 2
Apr 22 13:23:28 something.edu indico-uwsgi[985]: current working directory: /
Apr 22 13:23:28 something.edu indico-uwsgi[985]: detected binary path: /opt/indico/.venv/bin/uwsgi
Apr 22 13:23:28 something.edu indico-uwsgi[985]: your processes number limit is 15035
Apr 22 13:23:28 something.edu indico-uwsgi[985]: your memory page size is 4096 bytes
Apr 22 13:23:28 something.edu indico-uwsgi[985]: *** WARNING: you have enabled harakiri without post buffering. Slow upload could be rejected on post-unbuffered webservers ***
Apr 22 13:23:28 something.edu indico-uwsgi[985]: detected max file descriptor number: 1024
Apr 22 13:23:28 something.edu indico-uwsgi[985]: lock engine: pthread robust mutexes
Apr 22 13:23:28 something.edu indico-uwsgi[985]: thunder lock: disabled (you can enable it with --thunder-lock)
Apr 22 13:23:28 something.edu indico-uwsgi[985]: uwsgi socket 0 inherited UNIX address /opt/indico/web/uwsgi.sock fd 6
Apr 22 13:23:28 something.edu indico-uwsgi[985]: Python version: 3.12.12 (main, Dec 18 2025, 12:25:10) [GCC 11.4.0]
Apr 22 13:23:28 something.edu indico-uwsgi[985]: PEP 405 virtualenv detected: /opt/indico/.venv
Apr 22 13:23:28 something.edu indico-uwsgi[985]: Set PythonHome to /opt/indico/.venv
Apr 22 13:23:28 something.edu indico-uwsgi[985]: Python main interpreter initialized at 0x7f61e9f83dd0
Apr 22 13:23:28 something.edu indico-uwsgi[985]: python threads support enabled
Apr 22 13:23:28 something.edu indico-uwsgi[985]: your server socket listen backlog is limited to 100 connections
Apr 22 13:23:28 something.edu indico-uwsgi[985]: your mercy for graceful operations on workers is 60 seconds
Apr 22 13:23:28 something.edu indico-uwsgi[985]: mapped 446520 bytes (436 KB) for 4 cores
Apr 22 13:23:28 something.edu indico-uwsgi[985]: *** Operational MODE: preforking ***
Apr 22 13:28:30 something.edu indico-uwsgi[985]: /opt/indico/.venv/lib/python3.12/site-packages/sentry_sdk/_compat.py:201: Warning: IMPORTANT: We detected the use of uWSGI in preforking mode without thread support. This might lead to crashing workers. Please run uWSGI with both “–enable-threads” and “–py-call-uwsgi-fork-hooks” for full support.
Apr 22 13:28:30 something.edu indico-uwsgi[985]: warn(
Apr 22 13:28:34 something.edu indico-uwsgi[985]: WSGI app 0 (mountpoint=‘’) ready in 306 seconds on interpreter 0x7f61e9f83dd0 pid: 985 (default app)
Apr 22 13:28:34 something.edu indico-uwsgi[985]: gracefully (RE)spawned uWSGI master process (pid: 985)
Apr 22 13:28:34 something.edu indico-uwsgi[985]: sendmsg(): Transport endpoint is already connected [core/notify.c line 39]
Apr 22 13:28:34 something.edu indico-uwsgi[985]: spawned uWSGI worker 1 (pid: 3612851, cores: 1)
Apr 22 13:28:34 something.edu indico-uwsgi[985]: spawned uWSGI worker 2 (pid: 3612853, cores: 1)
Apr 22 13:28:34 something.edu indico-uwsgi[985]: spawned uWSGI worker 3 (pid: 3612854, cores: 1)
Apr 22 13:28:34 something.edu indico-uwsgi[3612851]: Wed Apr 22 13:28:34 2026 - mem-collector thread started for worker 1
Apr 22 13:28:34 something.edu indico-uwsgi[985]: spawned uWSGI worker 4 (pid: 3612855, cores: 1)
Apr 22 13:28:34 something.edu indico-uwsgi[985]: *** Stats server enabled on /opt/indico/web/uwsgi-stats.sock fd: 17 ***
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 20 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 18 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 224 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 3 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - error managing signal 0 on worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - uWSGI worker 2 error: the master disconnected from this worker. Shutting down the worker.
Apr 22 13:28:34 something.edu indico-uwsgi[3612854]: Wed Apr 22 13:28:34 2026 - mem-collector thread started for worker 3
Apr 22 13:28:34 something.edu indico-uwsgi[3612853]: Wed Apr 22 13:28:34 2026 - mem-collector thread started for worker 2
Apr 22 13:28:34 something.edu indico-uwsgi[3612855]: Wed Apr 22 13:28:34 2026 - mem-collector thread started for worker 4
Apr 22 13:28:36 something.edu indico-uwsgi[985]: DAMN ! worker 2 (pid: 3612853) died :(: trying respawn …
WSGI app 0 (mountpoint=‘’) ready in 306 seconds
So somehow it took 5 minutes to initialize. I have no idea why though…
PS: Please post such log output as code blocks, it’s much more readable that way.
Based on strace it appears to hang with a 5 minute timeout doing some sort of getpwuid but that is as far as I’ve made it. Probably some sort of (firewall?) misconfiguration here if I had to make a first guess.
Using uWSGI I have the same ~5s startup:
Apr 28 15:36:38 events uwsgi[274902]: *** Operational MODE: preforking+threaded ***
Apr 28 15:36:43 events uwsgi[274902]: WSGI app 0 (mountpoint='') ready in 5 seconds on interpreter 0x7f9924129610 pid: 274902 (default app)
You only have to find out on what’s defaulting to 300s before timing out. 