efab1dadde
Remove DEBUG_CACHES
2017-04-25 10:54:09 +01:00
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
d134d0935e
Only intern ascii strings
2017-04-24 14:07:48 +01:00
4d17add8de
Remove unused instance variable
2017-03-31 09:38:27 +01:00
6194a64ae9
Doc new instance variables
2017-03-30 14:19:10 +01:00
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
86780a8bc3
Don't convert to deferreds when not necessary
2017-03-30 14:14:36 +01:00
f9b4bb05e0
Fix the logcontext handling in the cache wrappers ( #2077 )
...
The cache wrappers had a habit of leaking the logcontext into the reactor while
the lookup function was running, and then not restoring it correctly when the
lookup function had completed. It's all the fault of
`preserve_context_over_{fn,deferred}` which are basically a bit broken.
2017-03-30 13:22:24 +01:00
95f21c7a66
Fix caching of remote servers' signature keys
...
The `@cached` decorator on `KeyStore._get_server_verify_key` was missing
its `num_args` parameter, which meant that it was returning the wrong key for
any server which had more than one recorded key.
By way of a fix, change the default for `num_args` to be *all* arguments. To
implement that, factor out a common base class for `CacheDescriptor` and `CacheListDescriptor`.
2017-03-22 15:11:30 +00:00
29ed09e80a
Fix assertion to stop transaction queue getting wedged
...
... and update some docstrings to correctly reflect the types being used.
get_new_device_msgs_for_remote can return a long under some circumstances,
which was being stored in last_device_list_stream_id_by_dest, and was then
upsetting things on the next loop.
2017-03-15 12:16:55 +00:00
3545e17f43
Add setdefault key to ExpiringCache
2017-03-10 10:30:49 +00:00
6b61060b51
Comment
2017-02-02 14:47:15 +00:00
9efcc3f3be
Comment
2017-02-02 13:50:22 +00:00
c430111d0e
Update LruCache size estimate on clear
2017-01-18 14:55:23 +00:00
380dba1020
Measure metrics of string_cache
2017-01-17 17:04:46 +00:00
37b4c7d8a9
Fix typo in return type
2017-01-17 14:43:32 +00:00
d6c75cb7c2
Rename and comment tree_to_leaves_iterator
2017-01-17 11:47:03 +00:00
1ccd5676e3
Remove needless call to evict()
2017-01-17 11:42:26 +00:00
f85b6ca494
Speed up cache size calculation
...
Instead of calculating the size of the cache repeatedly, which can take
a long time now that it can use a callback, instead cache the size and
update that on insertion and deletion.
This requires changing the cache descriptors to have two caches, one for
pending deferreds and the other for the actual values. There's no reason
to evict from the pending deferreds as they won't take up any more
memory.
2017-01-17 11:18:13 +00:00
6d00213e80
Use OrderedDict in ExpiringCache
2017-01-16 15:33:22 +00:00
46aebbbcbf
Add support for 'iterable' to ExpiringCache
2017-01-16 14:57:23 +00:00
2fae34bd2c
Optionally measure size of cache by sum of length of values
2017-01-13 17:46:17 +00:00
955f34d23e
Change get_pos_of_last_change to return upper bound
2016-09-15 15:12:07 +01:00
cb3edec6af
Use stream_change cache to make get_forward_extremeties_for_room cache more effective
2016-09-15 14:28:13 +01:00
45fd2c8942
Ensure invalidation list does not grow unboundedly
2016-08-19 16:09:16 +01:00
c0d7d9d642
Rename to on_invalidate
2016-08-19 15:13:58 +01:00
dc76a3e909
Make cache_context an explicit option
2016-08-19 15:02:38 +01:00
ba214a5e32
Remove lru option
2016-08-19 14:17:11 +01:00
4161ff2fc4
Add concept of cache contexts
2016-08-19 14:17:07 +01:00
248e6770ca
Cache federation state responses
2016-07-21 10:30:12 +01:00
3b096c5f5c
Merge branch 'erikj/cache_perf' of github.com:matrix-org/synapse into develop
2016-06-03 12:00:33 +01:00
58a224a651
Pull out update_results_dict
2016-06-03 11:47:07 +01:00
73c7112433
Change CacheMetrics to be quicker
...
We change it so that each cache has an individual CacheMetric, instead
of having one global CacheMetric. This means that when a cache tries to
increment a counter it does not need to go through so many indirections.
2016-06-03 11:26:52 +01:00
e043ede4a2
Small optimisation to CacheListDescriptor
2016-06-03 11:19:22 +01:00
597013caa5
Make cachedList go a bit faster
2016-06-03 11:13:29 +01:00
af03ecf352
Deduplicate joins
2016-04-07 14:19:02 +01:00
87f2dec8d4
Make the cache objects be per instance rather than being global
2016-04-06 13:08:05 +01:00
77cba688ed
Fix typo
2016-03-24 18:02:37 +00:00
191c7bef6b
Deduplicate identical /sync requests
2016-03-24 17:47:31 +00:00
8122ad7bab
Simplify intern_dict
2016-03-23 16:41:54 +00:00
acdfef7b14
Intern all the things
2016-03-23 16:25:54 +00:00
75daede92f
String intern
2016-03-23 14:53:53 +00:00
c4a8cbd15a
Make LruCache use a dedicated _Node class
2016-03-22 16:06:21 +00:00
a547e2df85
Return list, not generator.
2016-03-14 15:30:19 +00:00
239badea9b
Use syntax that works on both py2.7 and py3
2016-03-07 20:13:10 +00:00
374f9b2f07
Limit stream change cache size too
2016-03-01 13:30:15 +00:00
ce2cdced61
Move cache size fiddling to descriptors only. Fix tests
2016-03-01 13:21:46 +00:00
910fc0f28f
Add enviroment variable SYNAPSE_CACHE_FACTOR, default it to 0.1
2016-03-01 12:56:39 +00:00
72165e5b77
Reraise exception
2016-03-01 11:00:10 +00:00
ff2d7551c7
Correct cache miss detection
2016-03-01 10:59:17 +00:00
278d6c0527
Report size of ExpiringCache
2016-02-23 16:46:21 +00:00
c77dae7a1a
Change the way we figure out presence updates for small deltas
2016-02-23 14:54:40 +00:00
2c1fbea531
Fix up logcontexts
2016-02-08 14:26:45 +00:00
d83d004ccd
Fix flake8 warnings for new flake8
2016-02-02 17:18:50 +00:00
e70165039c
If stream pos is greater then earliest known key and entity hasn't changed, then entity hasn't changed
2016-01-29 16:41:32 +00:00
18579534ea
Prefill stream change caches
2016-01-29 14:37:59 +00:00
b18114e19e
Merge pull request #536 from matrix-org/erikj/sync
...
Make /sync "better".
2016-01-29 13:04:51 +00:00
fb7299800f
Directly set self.value
2016-01-29 11:29:14 +00:00
c046630c33
Remove spurious self.size
2016-01-29 11:17:54 +00:00
a30364c1f9
Correctly bookkeep the size of TreeCache
2016-01-29 10:44:46 +00:00
766526e114
Make TreeCache keep track of its own size.
2016-01-29 10:11:21 +00:00
50e18938a9
Reset size on clear
2016-01-29 10:00:45 +00:00
3f5dd18bd4
If the same as the earliest key, assume nothing has changed.
2016-01-28 18:11:41 +00:00
40431251cb
Correctly update _entity_to_key
2016-01-28 18:05:43 +00:00
82cf3a8043
Fix inequalities
2016-01-28 17:44:04 +00:00
0663c5bd52
Include cache hits with has_entity_changed
2016-01-28 17:27:28 +00:00
45cf827c8f
Change name and doc has_entity_changed
2016-01-28 16:39:18 +00:00
00cb3eb24b
Cache tags and account data
2016-01-28 16:37:41 +00:00
c23a8c7833
Ensure keys to RoomStreamChangeCache are ints
2016-01-28 15:55:26 +00:00
e1941442d4
Invalidate caches properly. Remove unused arg
2016-01-28 15:02:41 +00:00
b97f6626b6
Add cache to room stream
2016-01-27 17:33:26 +00:00
7cd418d38e
Don't add the member functiopn if we're not using treecache
2016-01-22 13:40:37 +00:00
cd80019eec
docs
2016-01-22 12:21:13 +00:00
d552861346
Revert all the bits changing keys of eeverything that used LRUCaches to tuples
2016-01-22 12:18:14 +00:00
10f76dc5da
Make LRU cache not default to treecache & add options to use it
2016-01-22 12:10:33 +00:00
5b142788d2
Add __contains__
2016-01-22 11:49:59 +00:00
eaa836e8ca
Docs for treecache
2016-01-22 11:47:22 +00:00
8acc5cb60f
Add invalidate_many here too
2016-01-22 11:22:32 +00:00
330be18ec5
peppate
2016-01-21 19:17:32 +00:00
f1f8122120
Change LRUCache to be tree-based so we can delete subtrees.
2016-01-21 19:16:25 +00:00
6c28ac260c
copyrights
2016-01-07 04:26:29 +00:00
d12c00bdc3
Add some docstring explaining the snapshot cache does
2015-12-23 15:18:11 +00:00
7fa71e3267
Add a unit test for the snapshot cache
2015-12-23 11:48:03 +00:00
9ac417fa88
Add a cache for initialSync responses that expires after 5 minutes
2015-12-22 18:27:56 +00:00
8e254862f4
Don't assume @cachedList function returns keys for everything
2015-08-18 11:11:33 +01:00
cfa62007a3
Docstring
2015-08-12 16:42:46 +01:00
d7451e0f22
Merge branch 'develop' of github.com:matrix-org/synapse into erikj/dictionary_cache
2015-08-12 10:30:30 +01:00
4807616e16
Wire up the dictionarycache to the metrics
2015-08-12 10:13:35 +01:00
2df8dd9b37
Move all the caches into their own package, synapse.util.caches
2015-08-11 18:00:59 +01:00
Erik Johnston