Commit Graph

558 Commits

Author SHA1 Message Date
Tim Abbott 728b852b28 logrotate: Dramatically decrease maximum size of logs.
It's rare that there's value in having the log files get this big, and
these changes mean these log files should never consume more than a
few gigabytes.

And in particular, the server.log is far more important than the other
log files, and grows much faster, so we might as well spend most of
the space we are spending on that.

I estimate that the total size of log files from this is going to be
under 1-2GB, since 75MB (compressed size) * 10 (compressed logs) +
500MB (uncompressed size) = 1.25GB from server.log, and the rest is
negligible.

Fixes part of #5724.
2017-07-07 12:58:33 -07:00
Tim Abbott 6355f26b91 supervisor: Dramatically decrease default log file max sizes.
Most of these log files are useless except a few minutes after an
event happens, and the aggregate effect of the originals size limits
meant that Zulip's logs could consume many gigabytes of disk.

The new logging strategy should limit our usage from supervisor logs
to at most 3 Gigabytes:
* 20 * 3 = 60MB per queue worker => <1GB.
* 100 * 10 = 1GB for Django and Tornado logs.

Fixes part of #5724.
2017-07-07 12:30:24 -07:00
cursiv 2b7938b8ed nginx: Add missing HTTP request methods to header.
This will make life easier if trying to write a Zulip client inside a
browser using the API.
2017-07-06 14:38:07 -07:00
rht 940cf9db3b Run queue processors multithreaded in production if system memory <3.5GB.
While running queue processors multithreaded will limit the
performance available to very small systems, it's easy to fix that by
adding more RAM, and previously, Zulip didn't work on such systems at
all, so this is unambiguously an improvement there.

Fixes #32.
Fixes #34.

(Commit message expanded significantly by tabbott.)
2017-06-03 12:19:58 -07:00
Aditya Bansal aa433f4342 pep8: Add compliance with rule E261 check_send_receive_time. 2017-05-31 17:07:15 -07:00
Aditya Bansal 5989c88545 pep8: Make compliant zulip-ec2-configure-interfaces with rule E261. 2017-05-31 17:07:15 -07:00
Aditya Bansal 6b8e85e065 pep8: Make compliant check_zephyr_mirror with rule E261. 2017-05-31 17:07:15 -07:00
Aditya Bansal 49ae51f23a pep8: Make compliant check_user_zephyr_mirror_liveness with rule E261. 2017-05-31 17:07:15 -07:00
Aditya Bansal 6d0927ed0b pep8: Add compliance with rule E261 to check_personal_zephyr_mirrors. 2017-05-31 17:07:15 -07:00
Tim Abbott c5bc1265f3 bots: Move log2zulip into api/integrations. 2017-05-26 15:15:56 -07:00
Tim Abbott 55f69c677a puppet: Remove obsolete zuliprc.nagios file.
This hasn't done anything for years.
2017-05-26 15:14:12 -07:00
Reid Barton ccb4c5c26f bots: Move zephyr-related files to api/integrations/zephyr/. 2017-05-26 15:07:02 -07:00
Elliott Jin 0ec9e54954 bots: Add queue and QueueProcessingWorker for embedded bots. 2017-05-25 15:00:51 -07:00
Konstantin Gukov dd76222a3f Fetch system bots using new get_system_bot function.
This eliminate a bunch of uninteresting calls to
get_user_profile_by_email.
2017-05-23 10:30:40 -07:00
Tim Abbott 1d1b0894a3 zulip_ops: Add logrotate configuration from main zulip. 2017-05-15 21:54:35 -07:00
Tim Abbott e049ea01b1 puppet: Update munin configuration to work with modern munin. 2017-05-15 21:49:53 -07:00
Tim Abbott 6871c227cd puppet: Add missing restarts of Nagios on config updates. 2017-05-15 21:49:53 -07:00
Tim Abbott e94586adb4 puppet: Fix path to events_deliver_enqueued_emails.log. 2017-05-09 22:29:58 -07:00
Aditya Bansal b677689fea pep8: Add compliance with rule E261 to pg_backup_and_purge.py. 2017-05-07 23:21:50 -07:00
vaibhav 8881b5eb9f Outgoing Webhook System: Check for @-mentioned outgoing webhook bots.
Also puts them into a processing queue, though the queue processor
does nothing.

Rewritten by tabbott to avoid unnecessary database queries in
do_send_messages.
2017-05-02 09:22:04 -07:00
K.Kanakhin e3e52e7284 email-mirror: Move postfix email mirror integration to separate script.
This fixes a performance problem where we were previously starting up
a full Django process (~0.7s even on a fast machine) every time a new
email came in, potentially allowing users to accidentally DoS a Zulip
server.  Now, we just post over HTTPS, allowing the existing thread
pool support to do its job.

- Add script wrapper to communicate postfix pipe with django web server
  over HTTP(S). It uses shared_secret authentication mode.
- Add django view to process messages from email mirror server.
- Clean management command `email-mirror`. Left just functional
  for cron email processing.
- Add routes for new tornado view.
- Change pipe script in master process postfix config template
  based on updated script.
- Add tests.

Tweaked by tabbott to adjust the directory and set better defaults.

Fixes #2421.
2017-04-24 21:24:23 -07:00
hackerkid b2504084ab Replace timezone.now with timezone_now. 2017-04-16 12:28:56 -07:00
Alexander Trost 0fb6779899 docker: Make user and dbname configurable in process_fts_updates. 2017-03-23 14:28:21 -07:00
Tim Abbott 17c8527856 nagios: Clean up check_send_receive_time arguments. 2017-03-15 12:52:27 -07:00
Tim Abbott 0ff1f3d663 nagios: Error on heartbeat events in check_send_receive_time.
It was probably going to fail anyway if those show up, but this
produces a clearer failure mode.
2017-03-15 12:51:19 -07:00
Tim Abbott 6a5e98b77e puppet: Increase MaxStartups SSH configuration. 2017-03-08 22:28:16 -08:00
K.Kanakhin 1e441b8d7c uwsgi: Add master mode to the main uwsgi process.
- Enable `master` parameter for `uswgi` configuration.
  It allows cleaning leaked processes if the parent
  process is closed unexpectedly or with SIGKILL command.
  Child processes follow to the master and kill themselves
  after the main process.

Fixes #3855
2017-03-07 21:35:51 -08:00
K.Kanakhin 6a801db1c2 missed-emails-sending: Move email sending to separate queue worker.
- Add new 'missedmessage_email_senders' queue for sending missed messages emails.
- Add the new worker to process 'missedmessage_email_senders' queue.
- Split aggregation missed messages and sending missed messages email
  to separate queue workers.
- Adapt tests for sending missed emails to the new logic.

Fixes #2607
2017-03-07 20:08:40 -08:00
Tim Abbott 67219cf660 puppet: Use restart-server for weekly server restarts.
Using `supervisorctl restart all` carried longer downtime (since it
just restarts everything at the same moment) and was less under our
control; I'm not sure it had any advantages.
2017-03-05 11:36:10 -08:00
Tim Abbott ab2ce9d5e9 lint: Fix whitespace error with recent mypy annotations. 2017-03-04 15:41:11 -08:00
Tim Abbott 75e81253f2 mypy: Work around several new mypy bugs in 0.501. 2017-03-04 15:33:39 -08:00
Raghav Jajodia a3a03bd6a5 mypy: Added Dict, List and Set imports.
Fixed mypy errors associated with the upgrade.
2017-03-04 14:33:44 -08:00
Rishi Gupta 3d07ac0c49 Change timezone-naive datetimes to use timezone.now() where safe to do so.
Change timezone-naive datetimes to use timezone.now() in cases where there
is no change in behavior.
2017-03-01 22:54:28 -08:00
Rishi Gupta 0218422e96 Use time.time() instead of datetime.now() to measure elapsed time.
Both because it is more idiomatic and because we will soon start enforcing
that all datetimes in Zulip are timezone aware.
2017-03-01 22:54:28 -08:00
Tim Abbott aed3632cbb puppet: Convert remaining queue workers to _ style. 2017-02-19 16:18:37 -08:00
Tim Abbott 4b54307d94 puppet: Generate most of main supervisor config with template. 2017-02-19 16:18:37 -08:00
Tim Abbott ae09d55e46 puppet: Move zulip.conf to be a template file. 2017-02-19 16:18:37 -08:00
Kouhei Sutou a2d935a2ee puppet: Fix PostgreSQL user to create PGroonga extension
"root" user isn't a PostgreSQL administrator. "postgres" is a PostgreSQL
administrator.
2017-02-08 12:57:56 -08:00
Tim Abbott fa8045a484 puppet: Add websockets Nagios test to configuration.
Since browser clients send messages via websockets and not the API,
this is an important element in making sure mission-critical Zulip
functionality is working.
2017-02-08 11:13:19 -08:00
Tim Abbott 8db13d0bb9 check_send_receive_time: Use a different state file for websockets.
Otherwise, the two Nagios checks will fight over the same state file
if both are in use.
2017-02-08 11:13:19 -08:00
Tim Abbott ba5f454be5 puppet: Extract zulip::analytics.
I'm not altogether happy with this (a better solution would be
database-level locking), but I think it solves the immediate problem
of folks with 2 servers being very likely to run analytics on both of
them.
2017-02-07 12:29:15 -08:00
Tim Abbott 70388b17d2 puppet: Add missing dependency on ssl-cert. 2017-02-06 15:51:38 -08:00
umkay 76f3d02590 analytics: Add cron job to run analytics jobs.
This adds a cron job to update the Zulip analytics counts, complete
with locking etc.

Substantially tweaked by tabbott.
2017-02-01 17:02:46 -08:00
Tim Abbott 2fb51ff876 puppet: Use SIGINT to restart uwsgi.
This results in a brief service interruption (not a graceful restart),
but fixes a bug where on a `supervisorctl restart zulip-django`, we'd
end up leaking a bunch of uwsgi processes.

The mechanism was that sending SIGHUP to uwsgi was a command for it to
gracefully restart, so it'd start doing that (whereas supervisor
expected it to be dying)... and then supervisor would start up the new
uwsgi process group, resulting in 2 uwsgi process groups running.

This, in turn, led to a memory leak that could eventually result in
OOM kills.
2017-01-28 22:26:12 -08:00
Tim Abbott 36d54cf5ff Replace references to zulip.com/dist with zulip.org/dist.
Now that zulip.org has all the files to distribute, there's no reason
to still point to the soon-to-be-decommissioned zulip.com/dist.
2017-01-28 17:56:25 -08:00
Eitan Adler 0ce29d7ad6 Remove some some duplicate words in copy. 2017-01-23 23:15:04 -08:00
Tim Abbott 4e171ce787 lint: Clean up E126 PEP-8 rule. 2017-01-23 22:06:13 -08:00
Tim Abbott d6e38e2a5c lint: Clean up E123 PEP-8 rule. 2017-01-23 21:34:26 -08:00
Tim Abbott bde2da7dfd lint: clean up PEP-8 W391 rule. 2017-01-23 20:39:02 -08:00
Tim Abbott bbd853e208 puppet: Add redirect to https to zulip.org. 2017-01-22 21:52:50 -08:00