Commit Graph

70 Commits

Author SHA1 Message Date
Richard van der Hoff
f749607c91 Make on_started synchronous too
This brings it into line with on_new_notifications and on_new_receipts. It
requires a little bit of hoop-jumping in EmailPusher to load the throttle
params before the first loop.
2018-10-22 16:12:11 +01:00
Richard van der Hoff
e7a16c6210 Remove redundant run_as_background_process() from pusherpool
`on_new_notifications` and `on_new_receipts` in `HttpPusher` and `EmailPusher`
now always return synchronously, so we can remove the `defer.gatherResults` on
their results, and the `run_as_background_process` wrappers can be removed too
because the PusherPool methods will now complete quickly enough.
2018-10-22 16:12:11 +01:00
Richard van der Hoff
c7273c11bc Give pushers their own background logcontext
Each pusher has its own loop which runs for as long as it has work to do. This
should run in its own background thread with its own logcontext, as other
similar loops elsewhere in the system do - which means that CPU usage is
consistently attributed to that loop, rather than to whatever request happened
to start the loop.
2018-10-22 16:12:11 +01:00
Amber Brown
7c27c4d51c
merge (#3576) 2018-09-14 03:11:11 +10:00
Amber Brown
49af402019 run isort 2018-07-09 16:09:20 +10:00
Amber Brown
07cad26d65
Remove all global reactor imports & pass it around explicitly (#3424) 2018-06-25 14:08:28 +01:00
Amber Brown
85ba83eb51 fixes 2018-05-22 16:28:23 -05:00
Amber Brown
df9f72d9e5 replacing portions 2018-05-21 19:47:37 -05:00
Adrian Tschira
57b58e2174 make imports local
Signed-off-by: Adrian Tschira <nota@notafile.com>
2018-04-28 13:41:41 +02:00
Richard van der Hoff
9255a6cb17 Improve exception handling for background processes
There were a bunch of places where we fire off a process to happen in the
background, but don't have any exception handling on it - instead relying on
the unhandled error being logged when the relevent deferred gets
garbage-collected.

This is unsatisfactory for a number of reasons:
 - logging on garbage collection is best-effort and may happen some time after
   the error, if at all
 - it can be hard to figure out where the error actually happened.
 - it is logged as a scary CRITICAL error which (a) I always forget to grep for
   and (b) it's not really CRITICAL if a background process we don't care about
   fails.

So this is an attempt to add exception handling to everything we fire off into
the background.
2018-04-27 11:07:40 +01:00
Richard van der Hoff
03dd745fe2 Better logging when pushes fail 2018-01-29 15:49:06 +00:00
Richard van der Hoff
e051abd20b add appid/device_display_name to to pusher logging 2018-01-29 15:04:16 +00:00
Richard van der Hoff
4528dd2443 Fix logging and add user_id 2018-01-22 20:15:42 +00:00
Richard van der Hoff
93efd7eb04 logging and debug for http pusher 2018-01-22 18:14:10 +00:00
David Baker
2a98ba0ed3 Rename redact_content option to include_content
The redact_content option never worked because it read the wrong config
section. The PR introducing it
(https://github.com/matrix-org/synapse/pull/2301) had feedback suggesting the
name be changed to not re-use the term 'redact' but this wasn't
incorporated.

This reanmes the option to give it a less confusing name, and also
means that people who've set the redact_content option won't suddenly
see a behaviour change when upgrading synapse, but instead can set
include_content if they want to.

This PR also updates the wording of the config comment to clarify
that this has no effect on event_id_only push.

Includes https://github.com/matrix-org/synapse/pull/2422
2017-11-08 10:35:30 +00:00
Richard van der Hoff
eaaabc6c4f replace 'except:' with 'except Exception:'
what could possibly go wrong
2017-10-23 15:52:32 +01:00
David Baker
436ee0a2ea Also include the room_id
as really it's part of the event ID
2017-09-18 15:58:38 +01:00
David Baker
b393f5db51 Use .get - it's much shorter 2017-09-18 15:50:26 +01:00
David Baker
a2562f9d74 Add support for event_id_only push format
Param in the data dict of a pusher that tells an HTTP pusher to
send just the event_id of the event it's notifying about and the
notification counts. For clients that want to go & fetch the body
of the event themselves anyway.
2017-09-18 15:39:39 +01:00
Caleb James DeLisle
75eba3b07d Fix TravisCI tests for PR #2301 2017-06-23 15:15:18 +02:00
Caleb James DeLisle
1591eddaea Add configuration parameter to allow redaction of content from push messages for google/apple devices 2017-06-23 13:01:04 +02:00
Erik Johnston
7356d52e73 Fix up push to use get_current_state_ids 2016-08-25 18:35:49 +01:00
Erik Johnston
b260f92936 Ignore AlreadyCalled errors on timer cancel 2016-07-31 16:00:12 +01:00
Mark Haines
0a7d3cd00f Create separate methods for getting messages to push
for the email and http pushers rather than trying to make a single
method that will work with their conflicting requirements.

The http pusher needs to get the messages in ascending stream order, and
doesn't want to miss a message.

The email pusher needs to get the messages in descending timestamp order,
and doesn't mind if it misses messages.
2016-07-28 20:24:24 +01:00
David Baker
870c45913e Use similar naming we use in email notifs for push
Fixes https://github.com/vector-im/vector-web/issues/1654
2016-06-24 11:41:11 +01:00
Mark Haines
a3ac837599 Optionally split out the pushers into a separate process 2016-04-21 17:22:37 +01:00
Erik Johnston
86e9bbc74e Add missing yield 2016-04-14 11:56:52 +01:00
Erik Johnston
e40f25ebe1 Rename log context 2016-04-14 11:54:14 +01:00
Erik Johnston
2ae91a9e2f Make send_badge private 2016-04-14 11:37:50 +01:00
Erik Johnston
56da835eaf Add necessary logging contexts 2016-04-14 11:33:50 +01:00
Erik Johnston
7be1065b8f Add extra Measure 2016-04-14 11:26:15 +01:00
David Baker
9bb041791c Run unsafe proces in a loop until we've caught up
and wrap unsafe process in a try block
2016-04-11 12:48:30 +01:00
David Baker
d96a070a3a Actually check if we;re processing 2016-04-08 16:49:39 +01:00
David Baker
7b6d519482 Make sure max stream ordering only increases 2016-04-08 14:08:16 +01:00
David Baker
52d1008661 Unsafe process should call itself if the max has changed 2016-04-08 14:06:54 +01:00
David Baker
ce3fe52498 Comment why unsafe process is unsafe 2016-04-08 14:02:38 +01:00
David Baker
a4a31fa8dc Only pass in what we need 2016-04-07 17:37:19 +01:00
David Baker
15e0f1696f Wrap process in a flag so we don't process whist already processing. 2016-04-07 17:31:08 +01:00
David Baker
fa129ce5b5 Add measure blocks 2016-04-07 17:12:29 +01:00
David Baker
e1e042f2a1 Add comments on min_stream_id
saying that the min stream id won't be completely accurate all the time
2016-04-07 17:09:36 +01:00
David Baker
d549fdfa22 Remove code that's now been obsoleted or moved elsewhere 2016-04-07 16:31:38 +01:00
David Baker
92e3071623 Send badge count pushes.
Also fix bugs with retrying.
2016-04-07 15:39:53 +01:00
David Baker
7e2c89a37f Make pushers use the event_push_actions table instead of listening on an event stream & running the rules again. Sytest passes, but remaining to do:
* Make badges work again
 * Remove old, unused code
2016-04-06 15:42:15 +01:00
Mark Haines
b9977ea667 Remove dead code for setting device specific rules.
It wasn't possible to hit the code from the API because of a typo
in parsing the request path. Since no-one was using the feature
we might as well remove the dead code.
2016-02-18 16:05:13 +00:00
David Baker
afb7b377f2 Merge branch 'develop' into push_badge_counts 2016-01-19 18:17:23 +00:00
David Baker
12623c99b6 Use the unread notification count to send accurate badge counts in push notifications. 2016-01-13 18:55:57 +00:00
Mark Haines
9c1f853d58 Rename 'user_name' to 'user_id' in push to make it consistent with the rest of the code 2016-01-13 13:32:59 +00:00
Matthew Hodgson
6c28ac260c copyrights 2016-01-07 04:26:29 +00:00
Mark Haines
37b2d69bbc Reuse a single http client, rather than creating new ones 2015-12-02 11:36:02 +00:00
David Baker
be9dafcd37 Dial down logging for failed pushers 2015-03-06 15:32:38 +00:00