Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Take value in a better way | Erik Johnston | 2016-09-05 | 1 | -1/+1 |
| | |||||
* | Correctly delete old state groups in purge history API | Erik Johnston | 2016-09-05 | 1 | -20/+79 |
| | |||||
* | Limit the length of state chains | Erik Johnston | 2016-09-02 | 1 | -21/+28 |
| | |||||
* | Generate state group ids in state layer | Erik Johnston | 2016-08-31 | 1 | -9/+1 |
| | |||||
* | Make _state_groups_id_gen a normal IdGenerator | Erik Johnston | 2016-08-30 | 1 | -44/+39 |
| | |||||
* | Use state handler instead of get_users_in_room/get_joined_hosts | Erik Johnston | 2016-08-26 | 1 | -1/+0 |
| | |||||
* | Preserve some logcontexts | Erik Johnston | 2016-08-24 | 1 | -6/+10 |
| | |||||
* | Move defer.returnValue out of Measure | Erik Johnston | 2016-08-19 | 1 | -1/+1 |
| | |||||
* | Measure _get_event_from_row | Erik Johnston | 2016-08-19 | 1 | -41/+43 |
| | |||||
* | Add missing database corruption recovery case | Benjamin Saunders | 2016-08-14 | 1 | -1/+2 |
| | | | | Signed-off-by: Benjamin Saunders <ben.e.saunders@gmail.com> | ||||
* | Raise 404 when couldn't find event | Erik Johnston | 2016-08-05 | 1 | -1/+1 |
| | |||||
* | Delete more tables | Erik Johnston | 2016-08-05 | 1 | -0/+5 |
| | |||||
* | Tweak integrity error recovery to work as intended | Benjamin Saunders | 2016-08-04 | 1 | -0/+3 |
| | |||||
* | Fix typo | Erik Johnston | 2016-08-04 | 1 | -3/+4 |
| | |||||
* | Retry event persistence on IntegrityError | Erik Johnston | 2016-08-04 | 1 | -4/+65 |
| | | | | | | | | | | Due to a bug in the porting script some backfilled events were not correctly persisted, causing irrecoverable IntegrityErrors on future attempts to persist those events. This commit adds a retry mechanism invoked upon IntegrityError, where when retried the tables are purged for all references to the events being persisted. | ||||
* | Remove other bit of deduplication | Erik Johnston | 2016-08-03 | 1 | -19/+0 |
| | |||||
* | Ensure we only persist an event once at a time | Erik Johnston | 2016-08-03 | 1 | -1/+18 |
| | |||||
* | Fix how push_actions are redacted. | Mark Haines | 2016-07-26 | 1 | -8/+10 |
| | |||||
* | Don't add rejected events if we've seen them befrore. Add some comments to ↵ | Mark Haines | 2016-07-26 | 1 | -5/+48 |
| | | | | explain what the code is doing mechanically | ||||
* | Only update the events and event_json tables for rejected events | Mark Haines | 2016-07-25 | 1 | -51/+62 |
| | |||||
* | Don't add the events to forward extremities if the event is rejected | Mark Haines | 2016-07-25 | 1 | -3/+6 |
| | |||||
* | Don't add rejections to the state_group, persist all rejections | Mark Haines | 2016-07-25 | 1 | -4/+5 |
| | |||||
* | Merge pull request #922 from matrix-org/erikj/file_api2 | Erik Johnston | 2016-07-20 | 1 | -0/+82 |
|\ | | | | | Feature: Add filter to /messages. Add 'contains_url' to filter. | ||||
| * | Add sender and contains_url field to events table | Erik Johnston | 2016-07-14 | 1 | -0/+82 |
| | | |||||
* | | Fix /purge_history bug | Erik Johnston | 2016-07-15 | 1 | -19/+15 |
| | | | | | | | | This was caused by trying to insert duplicate backward extremeties | ||||
* | | event_backwards_extremeties may not be empty | Erik Johnston | 2016-07-15 | 1 | -1/+11 |
|/ | |||||
* | Comment | Erik Johnston | 2016-07-07 | 1 | -0/+3 |
| | |||||
* | Return 400 rather than 500 | Erik Johnston | 2016-07-07 | 1 | -1/+4 |
| | |||||
* | Add rest servlet. Fix SQL. | Erik Johnston | 2016-07-06 | 1 | -5/+4 |
| | |||||
* | Add purge_history API | Erik Johnston | 2016-07-05 | 1 | -0/+6 |
| | |||||
* | Add storage function to purge history for a room | Erik Johnston | 2016-07-04 | 1 | -0/+140 |
| | |||||
* | Use similar naming we use in email notifs for push | David Baker | 2016-06-24 | 1 | -7/+0 |
| | | | | Fixes https://github.com/vector-im/vector-web/issues/1654 | ||||
* | Merge pull request #841 from matrix-org/erikj/event_counter | Erik Johnston | 2016-06-06 | 1 | -0/+9 |
|\ | | | | | Add metric counter for number of persisted events | ||||
| * | Add metric counter for number of persisted events | Erik Johnston | 2016-06-06 | 1 | -0/+9 |
| | | |||||
* | | Add events to cache when we persist them | Erik Johnston | 2016-06-06 | 1 | -0/+41 |
|/ | |||||
* | Don't rely on options when inserting event into cache | Erik Johnston | 2016-06-03 | 1 | -40/+43 |
| | |||||
* | Change the way we cache events | Erik Johnston | 2016-06-03 | 1 | -39/+41 |
| | |||||
* | Remove event fetching from DB threads | Erik Johnston | 2016-06-03 | 1 | -138/+0 |
| | |||||
* | Use state to calculate get_users_in_room | Erik Johnston | 2016-06-01 | 1 | -3/+0 |
| | |||||
* | Merge remote-tracking branch 'origin/develop' into dbkr/email_notifs | David Baker | 2016-05-10 | 1 | -11/+149 |
|\ | |||||
| * | Remove unused import | Erik Johnston | 2016-05-06 | 1 | -1/+1 |
| | | |||||
| * | Pull loop one level up | Erik Johnston | 2016-05-06 | 1 | -36/+41 |
| | | |||||
| * | Queue events for persistence | Erik Johnston | 2016-05-06 | 1 | -11/+144 |
| | | |||||
* | | Merge remote-tracking branch 'origin/develop' into dbkr/email_notifs | David Baker | 2016-04-29 | 1 | -2/+8 |
|\| | |||||
| * | Check that somethign has happend before running the selects | Mark Haines | 2016-04-27 | 1 | -2/+8 |
| | | |||||
* | | First bits of emailpusher | David Baker | 2016-04-19 | 1 | -0/+2 |
|/ | | | | Mostly logic of when to send an email | ||||
* | Add back backfilled parameter that was removed | Erik Johnston | 2016-04-12 | 1 | -3/+4 |
| | |||||
* | Check if we've already backfilled events | Erik Johnston | 2016-04-12 | 1 | -0/+16 |
| | |||||
* | Merge remote-tracking branch 'origin/develop' into ↵ | David Baker | 2016-04-07 | 1 | -2/+2 |
|\ | | | | | | | dbkr/pushers_use_event_actions | ||||
| * | Add a slaved events store class | Mark Haines | 2016-04-06 | 1 | -2/+2 |
| | | | | | | | | | | Add a test to check that get_room_names_and_aliases does the same thing on both the master and on the slave data store. | ||||
* | | pep8 | David Baker | 2016-04-06 | 1 | -1/+3 |
| | | |||||
* | | Make pushers use the event_push_actions table instead of listening on an ↵ | David Baker | 2016-04-06 | 1 | -0/+12 |
|/ | | | | | | | event stream & running the rules again. Sytest passes, but remaining to do: * Make badges work again * Remove old, unused code | ||||
* | Store invites in a separate table. | Erik Johnston | 2016-04-04 | 1 | -9/+4 |
| | |||||
* | Merge pull request #684 from matrix-org/markjh/backfill_id_gen | Mark Haines | 2016-04-01 | 1 | -14/+5 |
|\ | | | | | Use a stream id generator for backfilled ids | ||||
| * | Use a stream id generator for backfilled ids | Mark Haines | 2016-04-01 | 1 | -14/+5 |
| | | |||||
* | | Fix the invalidation of the names and aliases cache | Mark Haines | 2016-04-01 | 1 | -1/+1 |
|/ | |||||
* | Move the check for backfilled outside the for loop | Mark Haines | 2016-03-31 | 1 | -5/+5 |
| | |||||
* | Remove the is_new_state argument to persist event. | Mark Haines | 2016-03-31 | 1 | -40/+50 |
| | | | | | | | | | Move the checks for whether an event is new state inside persist event itself. This was harder than expected because there wasn't enough information passed to persist event to correctly handle invites from remote servers for new rooms. | ||||
* | Merge pull request #676 from matrix-org/markjh/replicate_stateIII | Mark Haines | 2016-03-31 | 1 | -2/+76 |
|\ | | | | | Add replication streams for ex outliers and current state resets | ||||
| * | Use a namedtuple rather than tuple unpacking | Mark Haines | 2016-03-31 | 1 | -2/+9 |
| | | |||||
| * | Add replication streams for ex outliers and current state resets | Mark Haines | 2016-03-30 | 1 | -1/+59 |
| | | |||||
| * | Add a entry to current_state_resets table when the current state is reset | Mark Haines | 2016-03-30 | 1 | -0/+9 |
| | | |||||
* | | return the state_group for backfill | Mark Haines | 2016-03-30 | 1 | -1/+2 |
|/ | |||||
* | Add a replication stream for state groups | Mark Haines | 2016-03-30 | 1 | -1/+5 |
| | |||||
* | Use a stream id generator to assign state group ids | Mark Haines | 2016-03-30 | 1 | -40/+50 |
| | |||||
* | Merge pull request #665 from matrix-org/erikj/dont_cache_events | Erik Johnston | 2016-03-23 | 1 | -2/+2 |
|\ | | | | | Only cache events in the event cache | ||||
| * | Don't cache events in get_current_state_for_key | Erik Johnston | 2016-03-23 | 1 | -2/+2 |
| | | |||||
* | | Make StateHandler._state_cache only store event_ids. | Erik Johnston | 2016-03-22 | 1 | -0/+25 |
|/ | |||||
* | Remove unused backfilled parameter from persist_event | Mark Haines | 2016-03-21 | 1 | -18/+4 |
| | |||||
* | Dedupe requested event list in _get_events | Erik Johnston | 2016-03-18 | 1 | -14/+12 |
| | |||||
* | Fix relative imports so they work in both py3 and py27 | Mark Haines | 2016-03-08 | 1 | -1/+1 |
| | |||||
* | Merge pull request #489 from matrix-org/markjh/replication | Mark Haines | 2016-03-01 | 1 | -0/+45 |
|\ | | | | | Add a /replication API for extracting the updates that happened on synapse. | ||||
| * | Add a /replication API for extracting the updates that happened on | Mark Haines | 2016-03-01 | 1 | -0/+45 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | synapse This is necessary for replicating the data in synapse to be visible to a separate service because presence and typing notifications aren't stored in a database so won't be visible to another process. This API can be used to either get the raw data by requesting the tables themselves or to just receive notifications for updates by following the streams meta-stream. Returns updates for each table requested a JSON array of arrays with a row for each row in the table. Each table is prefixed by a header row with the: name of the table, current stream_id position for the table, number of rows, number of columns and the names of the columns. This is followed by the rows that have been added to the server since the requester last asked. The API has a timeout and is hooked up to the notifier so that a slave can long poll for updates. | ||||
* | | Load the current id in the IdGenerator constructor | Mark Haines | 2016-03-01 | 1 | -3/+3 |
|/ | | | | | | | | | Rather than loading them lazily. This allows us to remove all the yield statements and spurious arguments for the get_next methods. It also allows us to replace all instances of get_next_txn with get_next since get_next no longer needs to access the db. | ||||
* | Remove unused param from get_max_token | Erik Johnston | 2016-02-18 | 1 | -1/+1 |
| | |||||
* | Rename functions | Erik Johnston | 2016-02-10 | 1 | -2/+2 |
| | |||||
* | Atomically persit push actions when we persist the event | Erik Johnston | 2016-02-09 | 1 | -10/+16 |
| | |||||
* | Fix up logcontexts | Erik Johnston | 2016-02-08 | 1 | -14/+20 |
| | |||||
* | Fix flake8 warnings for new flake8 | Daniel Wagner-Hall | 2016-02-02 | 1 | -3/+3 |
| | |||||
* | Cache tags and account data | Erik Johnston | 2016-01-28 | 1 | -1/+1 |
| | |||||
* | Invalidate caches properly. Remove unused arg | Erik Johnston | 2016-01-28 | 1 | -3/+6 |
| | |||||
* | PEP8 | Erik Johnston | 2016-01-27 | 1 | -1/+3 |
| | |||||
* | Use the same path for incremental with gap or without gap | Erik Johnston | 2016-01-27 | 1 | -1/+0 |
| | |||||
* | Add cache to room stream | Erik Johnston | 2016-01-27 | 1 | -0/+2 |
| | |||||
* | Add a Homeserver.setup method. | Erik Johnston | 2016-01-26 | 1 | -9/+5 |
| | | | | | | This is for setting up dependencies that require work on startup. This is useful for the DataStore that wants to read a bunch from the database before initiliazing. | ||||
* | copyrights | Matthew Hodgson | 2016-01-07 | 1 | -1/+1 |
| | |||||
* | Log when we skip daily messages | Daniel Wagner-Hall | 2016-01-06 | 1 | -1/+10 |
| | |||||
* | Track the time spent in the database per request. | Mark Haines | 2015-12-07 | 1 | -1/+1 |
| | | | | and track the number of transactions that request started. | ||||
* | Allow paginating search ordered by recents | Erik Johnston | 2015-11-30 | 1 | -0/+77 |
| | |||||
* | Implementation of state rollback in /sync | Richard van der Hoff | 2015-11-13 | 1 | -2/+4 |
| | | | | | | | Implementation of SPEC-254: roll back the state dictionary to how it looked at the start of the timeline. Merged PR https://github.com/matrix-org/synapse/pull/373 | ||||
* | Return world_readable and guest_can_join in /publicRooms | Daniel Wagner-Hall | 2015-11-10 | 1 | -0/+2 |
| | |||||
* | Open up /events to anonymous users for room events only | Daniel Wagner-Hall | 2015-11-05 | 1 | -0/+2 |
| | | | | Squash-merge of PR #345 from daniel/anonymousevents | ||||
* | Keep FTS indexes up to date. Only search through rooms currently joined | Erik Johnston | 2015-10-12 | 1 | -0/+2 |
| | |||||
* | Remove unused functions from synapse/storage/events.py | Mark Haines | 2015-09-23 | 1 | -11/+0 |
| | |||||
* | Add docstring | Daniel Wagner-Hall | 2015-09-22 | 1 | -0/+6 |
| | |||||
* | Implement configurable stats reporting | Daniel Wagner-Hall | 2015-09-22 | 1 | -1/+57 |
| | | | | | | | | | | SYN-287 This requires that HS owners either opt in or out of stats reporting. When --generate-config is passed, --report-stats must be specified If an already-generated config is used, and doesn't have the report_stats key, it is requested to be set. | ||||
* | Merge pull request #281 from matrix-org/erikj/update_extremeties | Erik Johnston | 2015-09-15 | 1 | -0/+2 |
|\ | | | | | When updating a stored event from outlier to non-outlier, remember to update the extremeties | ||||
| * | When updating a stored event from outlier to non-outlier, remember to update ↵ | Erik Johnston | 2015-09-15 | 1 | -0/+2 |
| | | | | | | | | the extremeties | ||||
* | | Fix indent | Mark Haines | 2015-08-24 | 1 | -1/+1 |
| | | |||||
* | | Remove syutil dependency in favour of smaller single-purpose libraries | Mark Haines | 2015-08-24 | 1 | -11/+12 |
|/ | |||||
* | Fix bug where we didn't correctly serialize the redacted_because key over ↵ | Erik Johnston | 2015-08-21 | 1 | -0/+2 |
| | | | | federation | ||||
* | Change Cache to not use *args in its interface | Erik Johnston | 2015-08-07 | 1 | -10/+11 |
| | |||||
* | Remove commented out code | Erik Johnston | 2015-07-10 | 1 | -14/+0 |
| | |||||
* | Add bulk insert events API | Erik Johnston | 2015-06-25 | 1 | -166/+249 |
| | |||||
* | Don't explode if we don't have the event | Erik Johnston | 2015-06-01 | 1 | -1/+3 |
| | |||||
* | Add config option to turn off freezing events. Use new encode_json api and ↵ | Erik Johnston | 2015-05-29 | 1 | -10/+13 |
| | | | | ujson.loads | ||||
* | Merge branch 'master' of github.com:matrix-org/synapse into develop | Erik Johnston | 2015-05-22 | 1 | -7/+18 |
|\ | |||||
| * | Add caches for things requested by the pushers | Mark Haines | 2015-05-21 | 1 | -7/+12 |
| | | |||||
| * | Add a cache for get_current_state with state_key | Mark Haines | 2015-05-21 | 1 | -0/+6 |
| | | |||||
* | | Merge pull request #155 from matrix-org/erikj/perf | Erik Johnston | 2015-05-21 | 1 | -51/+337 |
|\ \ | | | | | | | Bulk and batch retrieval of events. | ||||
| * | | s/for events/for requests for events/ | Erik Johnston | 2015-05-21 | 1 | -1/+1 |
| | | | |||||
| * | | s/for events/for requests for events/ | Erik Johnston | 2015-05-21 | 1 | -1/+1 |
| | | | |||||
| * | | Add comments | Erik Johnston | 2015-05-21 | 1 | -3/+14 |
| | | | |||||
| * | | Split up _get_events into defer and txn versions | Erik Johnston | 2015-05-20 | 1 | -24/+35 |
| | | | |||||
| * | | Make not return a deferred _get_event_from_row_txn | Erik Johnston | 2015-05-20 | 1 | -20/+5 |
| | | | |||||
| * | | PEP8 | Erik Johnston | 2015-05-20 | 1 | -2/+2 |
| | | | |||||
| * | | Split out _get_event_from_row back into defer and _txn version | Erik Johnston | 2015-05-20 | 1 | -11/+57 |
| | | | |||||
| * | | Comments and shuffle of functions | Erik Johnston | 2015-05-18 | 1 | -38/+42 |
| | | | |||||
| * | | Remove rejected events if we don't want rejected events | Erik Johnston | 2015-05-18 | 1 | -0/+3 |
| | | | |||||
| * | | PEP8 | Erik Johnston | 2015-05-15 | 1 | -1/+1 |
| | | | |||||
| * | | Revert limiting of fetching, it didn't help perf. | Erik Johnston | 2015-05-15 | 1 | -9/+2 |
| | | | |||||
| * | | init j | Erik Johnston | 2015-05-15 | 1 | -0/+1 |
| | | | |||||
| * | | Don't completely drain the list | Erik Johnston | 2015-05-15 | 1 | -2/+9 |
| | | | |||||
| * | | Make store.get_current_state fetch events asyncly | Erik Johnston | 2015-05-15 | 1 | -1/+0 |
| | | | |||||
| * | | Remove more debug logging | Erik Johnston | 2015-05-15 | 1 | -2/+0 |
| | | | |||||
| * | | Remove debug logging | Erik Johnston | 2015-05-15 | 1 | -11/+3 |
| | | | |||||
| * | | preserve log context | Erik Johnston | 2015-05-15 | 1 | -1/+2 |
| | | | |||||
| * | | Add a wait | Erik Johnston | 2015-05-15 | 1 | -2/+11 |
| | | | |||||
| * | | Srsly. Don't use closures. Baaaaaad | Erik Johnston | 2015-05-15 | 1 | -10/+8 |
| | | | |||||
| * | | Correctly pass through params | Erik Johnston | 2015-05-15 | 1 | -3/+3 |
| | | | |||||
| * | | Sort out error handling | Erik Johnston | 2015-05-15 | 1 | -25/+22 |
| | | | |||||
| * | | Fix daedlock | Erik Johnston | 2015-05-15 | 1 | -52/+73 |
| | | | |||||
| * | | Remove race condition | Erik Johnston | 2015-05-14 | 1 | -35/+46 |
| | | | |||||
| * | | Call from right thread | Erik Johnston | 2015-05-14 | 1 | -1/+3 |
| | | | |||||
| * | | Count and loop | Erik Johnston | 2015-05-14 | 1 | -36/+34 |
| | | | |||||
| * | | Don't start up more fetch_events | Erik Johnston | 2015-05-14 | 1 | -0/+1 |
| | | | |||||
| * | | Awful idea for speeding up fetching of events | Erik Johnston | 2015-05-14 | 1 | -36/+131 |
| | | | |||||
| * | | Turn off preemptive transactions | Erik Johnston | 2015-05-14 | 1 | -17/+7 |
| | | | |||||
| * | | PEP8 | Erik Johnston | 2015-05-14 | 1 | -2/+1 |
| | | | |||||
| * | | Move from _base to events | Erik Johnston | 2015-05-14 | 1 | -0/+246 |
| | | | |||||
* | | | Merge pull request #152 from matrix-org/notifier_performance | Mark Haines | 2015-05-18 | 1 | -0/+3 |
|\ \ \ | | | | | | | | | Notifier performance | ||||
| * | | | Merge branch 'notifier_unify' into notifier_performance | Mark Haines | 2015-05-14 | 1 | -0/+1 |
| |\| | | |||||
| * | | | Don't bother checking for updates if the stream token hasn't advanced for a user | Mark Haines | 2015-05-13 | 1 | -0/+3 |
| | | | | |||||
* | | | | Move get_events functions to storage.events | Erik Johnston | 2015-05-18 | 1 | -0/+132 |
| |/ / |/| | | |||||
* | | | You need to call contextmanager | Erik Johnston | 2015-05-12 | 1 | -0/+1 |
|/ / | |||||
* | | Merge pull request #149 from matrix-org/erikj/backfill | Mark Haines | 2015-05-12 | 1 | -6/+9 |
|\ \ | | | | | | | Backfill support | ||||
| * | | Do state groups persistence /after/ checking if we have already persisted ↵ | Erik Johnston | 2015-05-12 | 1 | -6/+9 |
| |/ | | | | | | | the event | ||||
* / | SYN-377: Make sure that the StreamIdGenerator.get_next.__exit__ is called ↵ | Mark Haines | 2015-05-12 | 1 | -19/+19 |
|/ | | | | from the main thread after the transaction completes, not from database thread before the transaction completes. | ||||
* | Merge branch 'erikj/executemany' of github.com:matrix-org/synapse into ↵ | Erik Johnston | 2015-05-05 | 1 | -11/+21 |
|\ | | | | | | | erikj/SYN-371 | ||||
| * | Merge branch 'develop' of github.com:matrix-org/synapse into erikj/executemany | Erik Johnston | 2015-05-05 | 1 | -11/+21 |
| |\ | |||||
| | * | Collect the invalidate callbacks on the transaction object rather than ↵ | Mark Haines | 2015-05-05 | 1 | -27/+21 |
| | | | | | | | | | | | | passing around a separate list | ||||
| | * | Invalidate the room_member cache if the current state events updates | Mark Haines | 2015-05-05 | 1 | -0/+5 |
| | | | |||||
| | * | Missing return for when the event was already persisted | Mark Haines | 2015-05-05 | 1 | -1/+1 |
| | | | |||||
| | * | Invalidate the caches from the correct thread | Mark Haines | 2015-05-05 | 1 | -13/+26 |
| | | | |||||
* | | | Don't insert without deduplication. In this case we never actually use this ↵ | Erik Johnston | 2015-05-05 | 1 | -22/+0 |
|/ / | | | | | | | table, so simply remove the insert entirely | ||||
* / | Add support for using executemany | Erik Johnston | 2015-05-05 | 1 | -20/+26 |
|/ | |||||
* | Shuffle operations so that locking upsert happens last in the txn. This ↵ | Erik Johnston | 2015-04-27 | 1 | -41/+41 |
| | | | | ensures the lock is held for the least amount of time possible. | ||||
* | Go back to storing JSON in TEXT | Erik Johnston | 2015-04-16 | 1 | -8/+8 |
| | |||||
* | Use True for True rather than 1 | Erik Johnston | 2015-04-15 | 1 | -5/+7 |
| | |||||
* | Add support for postgres instead of mysql. Change sql accourdingly. blob + ↵ | Erik Johnston | 2015-04-14 | 1 | -2/+2 |
| | | | | varbinary -> bytea. No support for UNSIGNED or CREATE INDEX IF NOT EXISTS. | ||||
* | Merge branch 'develop' of github.com:matrix-org/synapse into mysql | Erik Johnston | 2015-04-09 | 1 | -2/+2 |
|\ | |||||
| * | Indirect invalidations of _get_event_cache via a helper method to keep all ↵ | Paul "LeoNerd" Evans | 2015-03-25 | 1 | -2/+2 |
| | | | | | | | | uses of the cache lexically within one .py file | ||||
* | | Stream ordering and out of order insertions. | Erik Johnston | 2015-04-09 | 1 | -2/+7 |
| | | | | | | | | | | | | | | Handle the fact that events can be persisted out of order, and so to get the "current max" stream token becomes non trivial - as we need to make sure that *all* stream tokens less than the current max have also successfully been persisted. | ||||
* | | Don't use multiple UNIQUE constraints; it will cause deadlocks | Erik Johnston | 2015-04-07 | 1 | -7/+7 |
| | | |||||
* | | Don't use AUTOINCREMENT, use an in memory version | Erik Johnston | 2015-04-07 | 1 | -6/+7 |
| | | |||||
* | | Fix unicode database support | Erik Johnston | 2015-03-25 | 1 | -3/+5 |
| | | |||||
* | | Don't reinsert into event_edges | Erik Johnston | 2015-03-24 | 1 | -8/+8 |
| | | |||||
* | | Merge branch 'develop' of github.com:matrix-org/synapse into mysql | Erik Johnston | 2015-03-20 | 1 | -1/+1 |
|\| | |||||
| * | Merge branch 'develop' of github.com:matrix-org/synapse into store_rearrangement | Erik Johnston | 2015-03-20 | 1 | -0/+1 |
| | | |||||
| * | Rearrange storage modules | Erik Johnston | 2015-03-20 | 1 | -0/+394 |
| | |||||
* | Rearrange storage modules | Erik Johnston | 2015-03-20 | 1 | -0/+387 |