Commit Graph

451 Commits

Author SHA1 Message Date
Adrian Tschira
73cbdef5f7 fix py3 intern and remove unnecessary py3 encode
Signed-off-by: Adrian Tschira <nota@notafile.com>
2018-05-19 17:35:31 +02:00
Richard van der Hoff
093d8c415a Merge remote-tracking branch 'origin/develop' into rav/warn_on_logcontext_fail 2018-05-03 14:59:29 +01:00
Richard van der Hoff
a7fe62f0cb Fix logcontext leaks in rate limiter 2018-05-03 12:31:59 +01:00
Richard van der Hoff
415c6b672e Merge branch 'develop' into rav/more_logcontext_leaks 2018-05-02 16:16:01 +01:00
Richard van der Hoff
f22e7cda2c Fix a class of logcontext leaks
So, it turns out that if you have a first `Deferred` `D1`, you can add a
callback which returns another `Deferred` `D2`, and `D2` must then complete
before any further callbacks on `D1` will execute (and later callbacks on `D1`
get the *result* of `D2` rather than `D2` itself).

So, `D1` might have `called=True` (as in, it has started running its
callbacks), but any new callbacks added to `D1` won't get run until `D2`
completes - so if you `yield D1` in an `inlineCallbacks` function, your `yield`
will 'block'.

In conclusion: some of our assumptions in `logcontext` were invalid. We need to
make sure that we don't optimise out the logcontext juggling when this
situation happens. Fortunately, it is easy to detect by checking `D1.paused`.
2018-05-02 11:58:00 +01:00
Richard van der Hoff
e482f8cd85 Fix incorrect reference to StringIO
This was introduced in 4f2f5171
2018-05-02 09:12:26 +01:00
Richard van der Hoff
fdb6849b81
Merge pull request #3144 from matrix-org/rav/run_in_background_exception_handling
Trap exceptions thrown within run_in_background
2018-04-30 10:23:02 +01:00
Richard van der Hoff
db75c86e84
Merge branch 'develop' into py3-xrange-1 2018-04-30 01:02:25 +01:00
Richard van der Hoff
049b0b5af2
Merge pull request #3154 from NotAFile/py3-stringio
Replace stringIO imports with six
2018-04-30 00:59:04 +01:00
Richard van der Hoff
dbf6f28d64
Merge pull request #3155 from NotAFile/py3-bytes-1
more bytes strings
2018-04-30 00:38:21 +01:00
Richard van der Hoff
aab2e4da60
Merge pull request #3140 from matrix-org/rav/use_run_in_background
Use run_in_background in preference to preserve_fn
2018-04-30 00:34:28 +01:00
Adrian Tschira
e9143b6593 more bytes strings
Signed-off-by: Adrian Tschira <nota@notafile.com>
2018-04-29 00:13:57 +02:00
Adrian Tschira
d82b6ea9e6 Move more xrange to six
plus a bonus next()

Signed-off-by: Adrian Tschira <nota@notafile.com>
2018-04-28 13:57:00 +02:00
Adrian Tschira
4f2f5171b7 replace stringIO imports 2018-04-28 13:46:23 +02:00
Richard van der Hoff
fc149b4eeb Merge remote-tracking branch 'origin/develop' into rav/use_run_in_background 2018-04-27 14:31:23 +01:00
Richard van der Hoff
6146332387 Merge remote-tracking branch 'origin/develop' into rav/deferred_timeout 2018-04-27 14:18:00 +01:00
Richard van der Hoff
2a13af23bc Use run_in_background in preference to preserve_fn
While I was going through uses of preserve_fn for other PRs, I converted places
which only use the wrapped function once to use run_in_background, to avoid
creating the function object.
2018-04-27 12:55:51 +01:00
Richard van der Hoff
9d2c1b8429 Backport deferred.addTimeout
Twisted 16.0 doesn't have addTimeout, so let's backport it.
2018-04-27 12:52:30 +01:00
Richard van der Hoff
13843f771e Trap exceptions thrown within run_in_background
Turn any exceptions that get thrown synchronously within run_in_background into
Failures instead.
2018-04-27 12:17:13 +01:00
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
1ea904b9f0 Use deferred.addTimeout instead of time_bound_deferred
This doesn't feel like a wheel we need to reinvent.
2018-04-23 00:53:18 +01:00
Richard van der Hoff
8dc4a6144b
Merge pull request #3107 from NotAFile/py3-bool-nonzero
add __bool__ alias to __nonzero__ methods
2018-04-20 15:43:39 +01:00
Richard van der Hoff
c09a6daf09
Merge pull request #3110 from NotAFile/py3-six-queue
Replace Queue with six.moves.queue
2018-04-20 15:35:00 +01:00
Richard van der Hoff
11a67b7c9d
Merge pull request #3093 from matrix-org/rav/response_cache_wrap
Refactor ResponseCache usage
2018-04-20 11:31:17 +01:00
Adrian Tschira
878995e660 Replace Queue with six.moves.queue
and a six.range change which I missed the last time

Signed-off-by: Adrian Tschira <nota@notafile.com>
2018-04-16 00:46:21 +02:00
Adrian Tschira
f63ff73c7f add __bool__ alias to __nonzero__ methods
Signed-off-by: Adrian Tschira <nota@notafile.com>
2018-04-15 20:40:47 +02:00
Richard van der Hoff
d3347ad485 Revert "Use sortedcontainers instead of blist"
This reverts commit 9fbe70a7dc.

It turns out that sortedcontainers.SortedDict is not an exact match for
blist.sorteddict; in particular, `popitem()` removes things from the opposite
end of the dict.

This is trivial to fix, but I want to add some unit tests, and potentially some
more thought about it, before we do so.
2018-04-13 11:16:43 +01:00
Richard van der Hoff
60f6014bb7 ResponseCache: fix handling of completed results
Turns out that ObservableDeferred.observe doesn't return a deferred if the
result is already completed. Fix handling and improve documentation.
2018-04-13 07:32:29 +01:00
Richard van der Hoff
b78395b7fe Refactor ResponseCache usage
Adds a `.wrap` method to ResponseCache which wraps up the boilerplate of a
(get, set) pair, and then use it throughout the codebase.

This will be largely non-functional, but does include the following functional
changes:

* federation_server.on_context_state_request: drops use of _server_linearizer
  which looked redundant and could cause incorrect cache misses by yielding
  between the get and the set.
* RoomListHandler.get_remote_public_room_list(): fixes logcontext leaks
* the wrap function includes some logging. I'm hoping this won't be too noisy
  on production.
2018-04-12 13:02:15 +01:00
Richard van der Hoff
d5c74b9f6c
Merge pull request #3092 from matrix-org/rav/response_cache_metrics
Add metrics for ResponseCache
2018-04-12 12:59:36 +01:00
Richard van der Hoff
261124396e
Merge pull request #3059 from matrix-org/rav/doc_response_cache
Document the behaviour of ResponseCache
2018-04-12 11:22:30 +01:00
Richard van der Hoff
b3384232a0 Add metrics for ResponseCache 2018-04-10 23:14:47 +01:00
Vincent Breitmoser
9fbe70a7dc Use sortedcontainers instead of blist
This commit drop-in replaces blist with SortedContainers. They are
written in pure python so work with pypy, but perform as good as
native implementations, at least in a couple benchmarks:

http://www.grantjenks.com/docs/sortedcontainers/performance.html
2018-04-10 11:29:51 +02:00
Richard van der Hoff
13decdbf96 Revert "Merge pull request #3066 from matrix-org/rav/remove_redundant_metrics"
We aren't ready to release this yet, so I'm reverting it for now.

This reverts commit d1679a4ed7, reversing
changes made to e089100c62.
2018-04-09 12:59:12 +01:00
Richard van der Hoff
3449da3bc7
Merge pull request #3068 from matrix-org/rav/fix_cache_invalidation
Improve database cache performance
2018-04-05 17:21:44 +01:00
Richard van der Hoff
01afc563c3 Fix overzealous cache invalidation
Fixes an issue where a cache invalidation would invalidate *all* pending
entries, rather than just the entry that we intended to invalidate.
2018-04-05 16:24:04 +01:00
Richard van der Hoff
518f6de088 Remove redundant metrics which were deprecated in 0.27.0. 2018-04-04 19:46:28 +01:00
Richard van der Hoff
a9a74101a4 Document the behaviour of ResponseCache
it looks like everything that uses ResponseCache expects to have to
`make_deferred_yieldable` its results. It's debatable whether that is the best
approach, but let's document it for now to avoid further confusion.
2018-04-04 09:06:22 +01:00
Richard van der Hoff
05630758f2 Use static JSONEncoders
using json.dumps with custom options requires us to create a new JSONEncoder on
each call. It's more efficient to create one upfront and reuse it.
2018-03-29 23:13:33 +01:00
Matthew Hodgson
8cbbfaefc1 404 correctly on missing paths via NoResource
fixes https://github.com/matrix-org/synapse/issues/2043 and https://github.com/matrix-org/synapse/issues/2029
2018-03-23 10:32:50 +00:00
Erik Johnston
9a0d783c11 Add comments 2018-03-19 11:35:53 +00:00
Richard van der Hoff
5a6e54264d Make 'unexpected logging context' into warnings
I think we've now fixed enough of these that the rest can be logged at
warning.
2018-03-15 18:40:38 +00:00
Erik Johnston
7c7706f42b Fix bug where state cache used lots of memory
The state cache bases its size on the sum of the size of entries. The
size of the entry is calculated once on insertion, so it is important
that the size of entries does not change.

The DictionaryCache modified the entries size, which caused the state
cache to incorrectly think it was smaller than it actually was.
2018-03-15 15:46:54 +00:00
Richard van der Hoff
20f40348d4 Factor run_in_background out from preserve_fn
It annoys me that we create temporary function objects when there's really no
need for it. Let's factor the gubbins out of preserve_fn and start using it.
2018-03-08 11:50:11 +00:00
Richard van der Hoff
3a75de923b Rewrite make_deferred_yieldable avoiding inlineCallbacks
... because (a) it's actually simpler (b) it might be marginally more
performant?
2018-03-01 12:40:05 +00:00
Richard van der Hoff
bc496df192 report metrics on number of cache evictions 2018-02-05 15:34:01 +00:00
Matthew Hodgson
ab9f844aaf
Add federation_domain_whitelist option (#2820)
Add federation_domain_whitelist

gives a way to restrict which domains your HS is allowed to federate with.
useful mainly for gracefully preventing a private but internet-connected HS from trying to federate to the wider public Matrix network
2018-01-22 19:11:18 +01:00
Matthew Hodgson
d84f65255e
Merge pull request #2813 from matrix-org/matthew/registrations_require_3pid
add registrations_require_3pid and allow_local_3pids
2018-01-22 13:57:22 +00:00
Matthew Hodgson
8fe253f19b fix PR nitpicking 2018-01-19 18:23:45 +00:00
Matthew Hodgson
447f4f0d5f rewrite based on PR feedback:
* [ ] split config options into allowed_local_3pids and registrations_require_3pid
 * [ ] simplify and comment logic for picking registration flows
 * [ ] fix docstring and move check_3pid_allowed into a new util module
 * [ ] use check_3pid_allowed everywhere

@erikjohnston PTAL
2018-01-19 15:33:55 +00:00
Erik Johnston
b6dc7044a9
Merge pull request #2804 from matrix-org/erikj/file_consumer
Add decent impl of a FileConsumer
2018-01-18 16:31:33 +00:00
Richard van der Hoff
d57765fc8a Fix bugs in block metrics
... which I introduced in #2785
2018-01-18 12:24:42 +00:00
Erik Johnston
be0dfcd4a2 Do logcontexts correctly 2018-01-18 11:57:57 +00:00
Erik Johnston
1432f7ccd5 Move test stuff to tests 2018-01-18 11:57:57 +00:00
Erik Johnston
2f18a2647b Make all fields private 2018-01-18 11:57:54 +00:00
Erik Johnston
dc519602ac Ensure we registerProducer isn't called twice 2018-01-18 11:07:17 +00:00
Erik Johnston
17b54389fe Fix _notify_empty typo 2018-01-18 11:05:34 +00:00
Erik Johnston
28b338ed9b Move definition of paused_producer to __init__ 2018-01-18 11:04:41 +00:00
Erik Johnston
a177325b49 Fix comments 2018-01-18 11:02:43 +00:00
Erik Johnston
bc67e7d260 Add decent impl of a FileConsumer
Twisted core doesn't have a general purpose one, so we need to write one
ourselves.

Features:
- All writing happens in background thread
- Supports both push and pull producers
- Push producers get paused if the consumer falls behind
2018-01-17 16:43:03 +00:00
Richard van der Hoff
3d12d97415 Track DB scheduling delay per-request
For each request, track the amount of time spent waiting for a db
connection. This entails adding it to the LoggingContext and we may as well add
metrics for it while we are passing.
2018-01-16 17:23:32 +00:00
Richard van der Hoff
6324b65f08 Track db txn time in millisecs
... to reduce the amount of floating-point foo we do.
2018-01-16 15:53:18 +00:00
Richard van der Hoff
44a498418c Optimise LoggingContext creation and copying
It turns out that the only thing we use the __dict__ of LoggingContext for is
`request`, and given we create lots of LoggingContexts and then copy them every
time we do a db transaction or log line, using the __dict__ seems a bit
redundant. Let's try to optimise things by making the request attribute
explicit.
2018-01-16 15:49:42 +00:00
Richard van der Hoff
39f4e29d01 Reorganise request and block metrics
In order to circumvent the number of duplicate foo:count metrics increasing
without bounds, it's time for a rearrangement.

The following are all deprecated, and replaced with synapse_util_metrics_block_count:
  synapse_util_metrics_block_timer:count
  synapse_util_metrics_block_ru_utime:count
  synapse_util_metrics_block_ru_stime:count
  synapse_util_metrics_block_db_txn_count:count
  synapse_util_metrics_block_db_txn_duration:count

The following are all deprecated, and replaced with synapse_http_server_response_count:
   synapse_http_server_requests
   synapse_http_server_response_time:count
   synapse_http_server_response_ru_utime:count
   synapse_http_server_response_ru_stime:count
   synapse_http_server_response_db_txn_count:count
   synapse_http_server_response_db_txn_duration:count

The following are renamed (the old metrics are kept for now, but deprecated):

  synapse_util_metrics_block_timer:total ->
     synapse_util_metrics_block_time_seconds

  synapse_util_metrics_block_ru_utime:total ->
     synapse_util_metrics_block_ru_utime_seconds

  synapse_util_metrics_block_ru_stime:total ->
     synapse_util_metrics_block_ru_stime_seconds

  synapse_util_metrics_block_db_txn_count:total ->
     synapse_util_metrics_block_db_txn_count

  synapse_util_metrics_block_db_txn_duration:total ->
     synapse_util_metrics_block_db_txn_duration_seconds

  synapse_http_server_response_time:total ->
     synapse_http_server_response_time_seconds

  synapse_http_server_response_ru_utime:total ->
     synapse_http_server_response_ru_utime_seconds

  synapse_http_server_response_ru_stime:total ->
     synapse_http_server_response_ru_stime_seconds

   synapse_http_server_response_db_txn_count:total ->
      synapse_http_server_response_db_txn_count

   synapse_http_server_response_db_txn_duration:total
      synapse_http_server_response_db_txn_duration_seconds
2018-01-15 17:09:44 +00:00
Richard van der Hoff
b2cd6accf5 Remove __PreservingContextDeferred too 2017-11-14 23:00:10 +00:00
Richard van der Hoff
7e6fa29cb5 Remove preserve_context_over_{fn, deferred}
Both of these functions ae known to leak logcontexts. Replace the remaining
calls to them and kill them off.
2017-11-14 11:22:42 +00:00
Richard van der Hoff
bf993db11c Logging and logcontext fixes for Limiter
Add some logging to the Limiter in a similar spirit to the Linearizer, to help
debug issues.

Also fix a logcontext leak.

Also refactor slightly to avoid throwing exceptions.
2017-11-07 00:48:57 +00:00
Richard van der Hoff
0be99858f3 fix vars named l
E741 says "do not use variables named ‘l’, ‘O’, or ‘I’".
2017-10-23 15:56:38 +01:00
Richard van der Hoff
eaaabc6c4f replace 'except:' with 'except Exception:'
what could possibly go wrong
2017-10-23 15:52:32 +01:00
Richard van der Hoff
2e9f5ea31a Fix logcontext handling for persist_events
* don't use preserve_context_over_deferred, which is known broken.

* remove a redundant preserve_fn.

* add/improve some comments
2017-10-17 10:59:30 +01:00
Richard van der Hoff
cc794d60e7 Merge pull request #2532 from matrix-org/rav/fix_linearizer
Fix stackoverflow and logcontexts from linearizer
2017-10-11 17:29:32 +01:00
Richard van der Hoff
f30c4ed2bc logformatter: fix AttributeError
make sure we have the relevant fields before we try to log them.
2017-10-11 17:26:17 +01:00
Richard van der Hoff
4fad8efbfb Fix stackoverflow and logcontexts from linearizer
1. make it not blow out the stack when there are more than 50 things waiting
   for a lock. Fixes https://github.com/matrix-org/synapse/issues/2505.

2. Make it not mess up the log contexts.
2017-10-11 15:05:05 +01:00
Richard van der Hoff
3cc852d339 Fancy logformatter to format exceptions better
This is a bit of an experimental change at this point; the idea is to see if it
helps us track down where our stack overflows are coming from by logging the
stack when the exception was caught and turned into a Failure. (We'll also need
edf2704420).

If we deploy this, we'll be able to enable it via the log config yaml.
2017-10-09 17:44:42 +01:00
Richard van der Hoff
148428ce76 Fix logcontext handling for concurrently_execute
Avoid preserve_context_over_deferred, which is broken.
2017-10-06 22:24:28 +01:00
David Baker
8ad5f34908 pep8 2017-09-26 19:21:41 +01:00
David Baker
9fd086e506 unnecessary parens 2017-09-26 17:59:46 +01:00
David Baker
0b03a97708 Add module_loader.py 2017-09-26 17:56:41 +01:00
Erik Johnston
495f075b41 Increase default cache factor size. 2017-07-04 09:58:32 +01:00
Erik Johnston
b5e8d529e6 Define CACHE_SIZE_FACTOR once 2017-07-04 09:56:44 +01:00
Erik Johnston
c72058bcc6 Use an ExpiringCache for storing registration sessions
This is because pruning them was a significant performance drain on
matrix.org
2017-06-29 14:08:37 +01:00
Erik Johnston
efc2b7db95 Rewrite conditional 2017-06-09 13:35:15 +01:00
Erik Johnston
eed59dcc1e Fix has_any_entity_changed
Occaisonally has_any_entity_changed would throw the error: "Set changed
size during iteration" when taking the max of the `sorteddict`. While
its uncertain how that happens, its quite inefficient to iterate over
the entire dict anyway so we change to using the more traditional
`bisect_*` functions.
2017-06-09 11:44:01 +01:00
Erik Johnston
304880d185 Add stream change cache 2017-05-31 15:46:36 +01:00
Erik Johnston
bd7bb5df71 Pull out if statement from for loop 2017-05-22 15:12:19 +01:00
Erik Johnston
e3417a06e2 Update list cache to handle one arg case
We update the normal cache descriptors to handle caches with a single
argument specially so that the key wasn't a 1-tuple. We need to update
the cache list to be aware of this.
2017-05-22 15:04:42 +01:00
Erik Johnston
bbfe4e996c Make get_state_groups_from_groups faster.
Most of the time was spent copying a dict to filter out sentinel values
that indicated that keys did not exist in the dict. The sentinel values
were added to ensure that we cached the non-existence of keys.

By updating DictionaryCache to keep track of which keys were known to
not exist itself we can remove a dictionary copy.
2017-05-17 15:12:15 +01:00
Erik Johnston
ffad4fe35b Don't update event cache hit ratio from get_joined_users
Otherwise the hit ration of plain get_events gets completely skewed by
calls to get_joined_users* functions.
2017-05-08 16:06:17 +01:00
Erik Johnston
d2d8ed4884 Optimise caches with single key 2017-05-04 14:18:46 +01:00
Richard van der Hoff
2e996271fe Instantiate DeferredTimedOutError correctly
Call `super` correctly, so that we correctly initialise the `errcode` field.

Fixes https://github.com/matrix-org/synapse/issues/2179.
2017-05-02 13:26:17 +01:00
Erik Johnston
d9aa645f86 Reduce size of joined_user cache
The _get_joined_users_from_context cache stores a mapping from user_id
to avatar_url and display_name. Instead of storing those in a dict,
store them in a namedtuple as that uses much less memory.

We also try converting the string to ascii to further reduce the size.
2017-04-25 14:38:51 +01:00
Erik Johnston
efab1dadde Remove DEBUG_CACHES 2017-04-25 10:54:09 +01:00
Erik Johnston
119cb9bbcf Reduce cache size by not storing deferreds
Currently the cache descriptors store deferreds rather than raw values,
this is a simple way of triggering only one database hit and sharing the
result if two callers attempt to get the same value.

However, there are a few caches that simply store a mapping from string
to string (or int). These caches can have a large number of entries,
under the assumption that each entry is small. However, the size of a
deferred (specifically the size of ObservableDeferred) is signigicantly
larger than that of the raw value, 2kb vs 32b.

This PR therefore changes the cache descriptors to store the raw values
rather than the deferreds.

As a side effect cached storage function now either return a deferred or
the actual value, as the cached list decriptor already does. This is
fine as we always end up just yield'ing on the returned value
eventually, which handles that case correctly.
2017-04-25 10:23:11 +01:00
Erik Johnston
d134d0935e Only intern ascii strings 2017-04-24 14:07:48 +01:00
Richard van der Hoff
e2eebf1696 Fix fixme in preserve_fn
`preserve_fn` is no longer used as a decorator anywhere, so we can safely fix a
fixme therein.
2017-04-03 15:38:02 +01:00
Erik Johnston
4d17add8de Remove unused instance variable 2017-03-31 09:38:27 +01:00
Erik Johnston
5b5b171f3e Docs 2017-03-30 17:05:53 +01:00
Erik Johnston
b282fe7170 Revert log context change 2017-03-30 17:03:59 +01:00
Erik Johnston
6194a64ae9 Doc new instance variables 2017-03-30 14:19:10 +01:00
Erik Johnston
014fee93b3 Manually calculate cache key as getcallargs is expensive
This is because getcallargs recomputes the getargspec, amongst other
things, which we don't need to do as its already been done
2017-03-30 14:14:46 +01:00