mirror/bugsink
1
0
Fork 0
mirror of https://github.com/bugsink/bugsink.git synced 2025-12-30 18:00:17 -06:00
Code Issues Packages Projects Releases Wiki Activity
583 Commits 46 Branches 58 Tags
123bc41b93822a6b6a4d46bb7e787a7beab42c04
Commit Graph

583 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Klaas van Schelven
123bc41b93 User registration 2024-05-29 15:43:00 +02:00
Klaas van Schelven
cef1127e48 Make user-model swappable 
I may just need this later, and doing it this late was already painful enough.
 2024-05-29 10:22:57 +02:00
Klaas van Schelven
9b6409f0cd Non-ugly login page 2024-05-27 15:10:01 +02:00
Klaas van Schelven
7cf521ce74 Timed SQLite backend 2024-05-27 14:21:26 +02:00
Klaas van Schelven
9f1be0b001 Document admin dead-end 2024-05-27 13:54:02 +02:00
Klaas van Schelven
e6d27e881f Add process_django_templates to nohup.py 2024-05-27 13:54:02 +02:00
Klaas van Schelven
8a6e721dbd Update sentry SDK dependency 
better sooner than later given the extensions; cursory check and
reading the 'breaking changes' seems to indicate: all is well
 2024-05-27 13:54:02 +02:00
Klaas van Schelven
a6123e22ac Fail more loudly when sentry sdk experiences network troubles 2024-05-27 13:54:02 +02:00
Klaas van Schelven
4c48376b5f nohub implementations should never cause Exceptions themselves 2024-05-27 13:54:02 +02:00
Klaas van Schelven
29b69e51b8 Add CaptureExceptionsMiddleware 2024-05-27 13:54:02 +02:00
Klaas van Schelven
1be377f962 Explain why LoginRequiredMiddleware.process_view returning an HttpResponse is OK 2024-05-27 13:54:02 +02:00
Klaas van Schelven
c811ddb6fb Don't call the view yourself (and return a result) from process_view 
calling the view yourself short-circuits the correct handling, which is not
what we want. It's not abundantly clear from the docs that this is generally
not what we want, but it is cimplied:

> It should return either None or an HttpResponse object. If it returns None,
> Django will continue processing this request, executing any other
> process_view() middleware and, then, the appropriate view. If it returns an
> HttpResponse object, Django won’t bother calling the appropriate view; it’ll
> apply response middleware to that HttpResponse and return the result.

We don't want to do any short-circuiting here, we just want to measure time
and have a view_name to display. Moving stuff around fixes it.

(That this was an error became clear when I wanted to do process_exception
in another Middleware, and that was never called)
 2024-05-27 13:54:02 +02:00
Klaas van Schelven
049053cc51 sentry_sdk_extension: nohub 2024-05-27 13:54:02 +02:00
Klaas van Schelven
213d6d9a32 Sentry SDK troubles: document them 2024-05-27 13:54:02 +02:00
Klaas van Schelven
55d1aa463c Yet another version of the cornless business 2024-05-27 13:54:02 +02:00
Klaas van Schelven
b5fbcfa7c8 Add 2 versions of cornless that are useful for sqlite debugging 2024-05-27 13:54:02 +02:00
Klaas van Schelven
082986f8ad Cornless: an interruptable server 
does not work for sqlite though, because that c-code does not
regularly check whether signal handling is needed.
 2024-05-27 13:54:01 +02:00
Klaas van Schelven
e6403ccfe2 cornless util added 2024-05-23 15:34:13 +02:00
Klaas van Schelven
8b0a4aafe0 Note playground.bugsink findings of get_period_counter initialization 2024-05-23 14:32:30 +02:00
Klaas van Schelven
fb66b04be9 Document playground.bugsink.com performance findings 2024-05-23 14:24:04 +02:00
Klaas van Schelven
4647c1b498 Stress test: better stats 2024-05-23 12:36:14 +02:00
Klaas van Schelven
f0c255a346 Force envelope use in stress tests 2024-05-23 11:02:44 +02:00
Klaas van Schelven
75b620941a Don't (accidentally) load all events into memory on-init 2024-05-23 10:43:28 +02:00
Klaas van Schelven
82d40e3741 Push get_pc_registry into snappea.foreman init 2024-05-23 10:12:24 +02:00
Klaas van Schelven
093f81f2f2 Log 'period counter registry initialization' 2024-05-23 10:08:11 +02:00
Klaas van Schelven
23d7f2172d Add showstat (for snappea queue size) 2024-05-23 08:37:48 +02:00
Klaas van Schelven
aee60969ad Fix typo 2024-05-22 21:02:24 +02:00
Klaas van Schelven
a24c295535 Forgotten file added 2024-05-22 20:57:16 +02:00
Klaas van Schelven
78a85087a2 'performance' info for event bytes count 2024-05-22 20:43:10 +02:00
Klaas van Schelven
5af1d2384e Performance logging of Snappea task create/delete 2024-05-22 17:45:28 +02:00
Klaas van Schelven
34f58f9106 Logging fixes 
* only use snappea handler when there was already something configured
* in development, use the 'snappea' rather than the 'look below' version for snappea
 2024-05-22 17:34:29 +02:00
Klaas van Schelven
e331ef6d79 Clarify development loggers somewhat 2024-05-22 17:12:19 +02:00
Klaas van Schelven
151af98559 Performance logging: push into development.py (i.e. remove from non-development servers) 2024-05-22 17:06:21 +02:00
Klaas van Schelven
9f53ba5a8f Remove timestamps from default snappea 
e.g. systemd otherwise has double timestamps
 2024-05-22 13:08:24 +02:00
Klaas van Schelven
4624c1d88c Default max request size: also nginx-side 2024-05-22 13:02:51 +02:00
Klaas van Schelven
b09cfb21c3 Configure runsnappea to be rebooted every day 
a way to limit memory leaks

also: deal with SIGTERM 'correctly' (i.e. as demanded by systemd)
 2024-05-22 12:44:58 +02:00
Klaas van Schelven
9fcbfb9033 Add templates for 404, 500 2024-05-22 09:37:51 +02:00
Klaas van Schelven
7c760ba546 Remove RuntimeDirectory from runsnappea systemd example conf 2024-05-22 09:24:53 +02:00
Klaas van Schelven
529a7a7c9f runssnappea as a separate command (for e.g. ps) 2024-05-22 09:23:31 +02:00
Klaas van Schelven
bf0855b11d Gunicorn: max-requests 
just picked 'some reasonably seeming value' to avoid memory leaks (which I've not
observed yet, but are easily imagined)
 2024-05-22 08:43:57 +02:00
Klaas van Schelven
f150c839dc Recommended setup: fix tmpfile troubles in 2 ways 
* recommend to just run in the home dir
* don't use private tmp

The troubles were: when set up using private tmp files, the 2 processes
cannot communicate with each other
 2024-05-22 08:37:52 +02:00
Klaas van Schelven
d4c74c22f8 In the recommended setup, just use port 8000 
simplify simplify simplify. In particular, this creates a symmetry with a 'just run
manage.py runserver', which could be used for debugging

Also, turns out sockets aren't actually faster:
https://stackoverflow.com/a/54013893/339144

> In this particular setup no unix socket performance advantage.
 2024-05-22 08:30:16 +02:00
Klaas van Schelven
c92e3ec772 Add INGEST_STORE_BASE_DIR 2024-05-20 09:44:38 +02:00
Klaas van Schelven
653bbe0dd2 Remove 'default' conf; it doesn't really map to any particular idea we express in the docs 2024-05-20 09:27:50 +02:00
Klaas van Schelven
53353d2d8c Add some more percentiles 2024-05-19 15:36:27 +02:00
Klaas van Schelven
4b5f1371ed docs: Add server_name to the 443 server too 
this is needed when you have multiple servers on the same host
i.e. I ran into it when running munin on the same server
 2024-05-18 08:24:36 +02:00
Klaas van Schelven
89dba6e6e5 Fix typo 2024-05-17 15:52:43 +02:00
Klaas van Schelven
0d1f966914 Header configuration: settings-side to correspond with the recommended setup 2024-05-17 12:30:10 +02:00
Klaas van Schelven
371c86b552 Docs for single-server: cleanup by doing a full replay 2024-05-17 12:23:17 +02:00
Klaas van Schelven
5ff2623112 Add checksnappea command 2024-05-17 12:03:40 +02:00