summary refs log tree commit diff
path: root/scripts-dev/tail-synapse.py (unfollow)
Commit message (Collapse)AuthorFilesLines
2017-04-21Remove redundant try/except clausesRichard van der Hoff1-23/+11
The `except SynapseError` clauses were pointless because the wrapped functions would never throw a `SynapseError` (they either throw a `CodeMessageException` or a `RuntimeError`). The `except CodeMessageException` is now also pointless because the caller treats all exceptions equally, so we may as well just throw the `CodeMessageException`.
2017-04-21Broaden the conditions for locally_rejecting invitesRichard van der Hoff1-1/+7
The logic for marking invites as locally rejected was all well and good, but didn't happen when the remote server returned a 500, or wasn't reachable, or had no DNS, or whatever. Just expand the except clause to catch everything. Fixes https://github.com/matrix-org/synapse/issues/761.
2017-04-21Remove redundant functionRichard van der Hoff1-10/+5
inline `reject_remote_invite`, which only existed to make tracing the callflow more difficult.
2017-04-21Fix some lies, and other clarifications, in docstringsRichard van der Hoff4-19/+88
The documentation on get_json has been wrong ever since the very first commit to synapse...
2017-04-18m.read_marker -> m.fully_read (#2128)Luke Barnard3-9/+9
Also: - change the REST endpoint to have a "S" on the end (so it's now /read_markers) - change the content of the m.read_up_to event to have the key "event_id" instead of "marker".
2017-04-17Update README.rstMark Felder1-1/+1
The FreeBSD port has been moved to the net-im category
2017-04-13Correct logic in is_event_afterLuke Barnard1-1/+1
2017-04-12Simplify is_event_after logicLuke Barnard2-12/+6
2017-04-12More null-guard changesLuke Barnard1-4/+4
2017-04-12Remove Unused ref to hsLuke Barnard1-1/+0
2017-04-12Move a spaceLuke Barnard1-2/+2
2017-04-12Only notify user, not entire roomLuke Barnard1-3/+1
2017-04-12Remove comment, simplify null-guardLuke Barnard1-4/+1
2017-04-12Add some commentsErik Johnston1-0/+15
2017-04-12Remove last reference to worker_replication_urlErik Johnston1-1/+2
2017-04-12CommentErik Johnston1-2/+1
2017-04-12Reuse get_interested_partiesErik Johnston2-19/+8
2017-04-12Handle no previous RMLuke Barnard1-1/+4
2017-04-11travis flake8..Luke Barnard1-0/+1
2017-04-11flake8Luke Barnard1-1/+0
2017-04-11Refactor event ordering check to events storeLuke Barnard2-28/+32
2017-04-11CopyrightLuke Barnard1-1/+1
2017-04-11Add a counter metric for successfully-sent transactionsPaul "LeoNerd" Evans1-0/+3
2017-04-11flake8Luke Barnard3-5/+4
2017-04-11Bail on where clause insteadErik Johnston1-3/+4
2017-04-11Don't bother with outer check for nowErik Johnston1-3/+0
2017-04-11Update replication docsErik Johnston1-22/+4
2017-04-11Update workers docsErik Johnston1-10/+5
2017-04-11Move get_presence_list_* to SlaveStoreErik Johnston3-24/+13
2017-04-11CommentErik Johnston2-1/+9
2017-04-11Move get_interested_partiesErik Johnston2-38/+36
2017-04-11CommentsErik Johnston2-9/+10
2017-04-11Move get_interested_remotes back to presence handlerErik Johnston2-38/+55
2017-04-11CommentsErik Johnston1-1/+14
2017-04-11Finish implementing RM endpointLuke Barnard5-48/+67
- This change causes a 405 to be sent if "m.read_marker" is set via /account_data - This also fixes-up the RM endpoint so that it actually Works.
2017-04-11Initial commit of RM server-side impllukebarnard2-0/+153
(See https://docs.google.com/document/d/1UWqdS-e1sdwkLDUY0wA4gZyIkRp-ekjsLZ8k6g_Zvso/edit#heading=h.lndohpg8at5u)
2017-04-11Bump changelog github/release-v0.20.0 release-v0.20.0Erik Johnston1-1/+1
2017-04-11Remove test of replication resourceErik Johnston1-204/+0
2017-04-11Fix getting latest device IP for user with no devicesErik Johnston1-0/+5
2017-04-11Remove HTTP replication APIsErik Johnston5-753/+0
2017-04-10trust a hypothetical future riot.im ISMatthew Hodgson1-0/+1
2017-04-10Use generatorsErik Johnston1-2/+2
2017-04-10Dedupe KeyedEdu and Devices federation repl trafficErik Johnston1-6/+9
2017-04-10Reduce federation presence replication trafficErik Johnston6-80/+139
This is mainly done by moving the calculation of where to send presence updates from the presence handler to the transaction queue, so we only need to send the presence event (and not the destinations) across the replication connection. Before we were duplicating by sending the full state across once per destination.
2017-04-10TypoErik Johnston1-1/+1
2017-04-10Rename variableErik Johnston1-7/+7
2017-04-10Up replication ping timeoutErik Johnston1-2/+4
2017-04-10Revert to sending the same data type as beforeErik Johnston1-2/+2
2017-04-10Change nameErik Johnston1-2/+2
2017-04-10CommentsErik Johnston2-6/+8
2017-04-07When we do an invite rejection, save the signed leave event to the dbRichard van der Hoff1-4/+4
During a rejection of an invite received over federation, we ask a remote server to make us a `leave` event, then sign it, then send that with `send_leave`. We were saving the *unsigned* version of the event (which has a different event id to the signed version) to our db (and sending it to the clients), whereas other servers in the room will have seen the *signed* version. We're not aware of any actual problems that caused, except that it makes the database confusing to look at and generally leaves the room in a weird state.
2017-04-07CommentErik Johnston1-0/+9
2017-04-07CommentErik Johnston1-2/+15
2017-04-07Add loggingErik Johnston1-0/+8
2017-04-07Fix up federation SendQueue and document typesErik Johnston2-91/+221
2017-04-07Speed up get_current_state_idsErik Johnston1-10/+17
Using _simple_select_list is fairly expensive for functions that return a lot of rows and/or get called a lot. (This is because it carefully constructs a list of dicts). get_current_state_ids gets called a lot on startup and e.g. when the IRC bridge decided to send tonnes of joins/leaves (as it invalidates the cache). We therefore replace it with a custon txn function that builds up the final result dict without building up and intermediate representation.
2017-04-07Bump version and changelogErik Johnston2-1/+10
2017-04-06fix typo in synctl helpKim Brose1-1/+1
2017-04-06Use iteritemsErik Johnston1-2/+2
2017-04-06CommentErik Johnston1-1/+1
2017-04-06Document types of the replication streamsErik Johnston1-28/+76
2017-04-06DocsErik Johnston1-0/+7
2017-04-06Fix incorrect type when using InvalidateCacheCommandErik Johnston1-1/+1
2017-04-05Support authenticated SMTPDaniel Dent2-1/+21
Closes (SYN-714) #1385 Signed-off-by: Daniel Dent <matrixcontrib@contactdaniel.net>
2017-04-05Add log linesErik Johnston1-1/+2
2017-04-05Rearrange metricsErik Johnston2-19/+31
2017-04-05Fix typoErik Johnston1-2/+2
2017-04-05Fixup some metrics for tcp replErik Johnston2-0/+19
2017-04-05Don't immediately notify the master about users whose syncs have gone awayErik Johnston1-4/+36
2017-04-05Don't double json encode typing replication dataErik Johnston2-6/+2
2017-04-05Don't double json encode federation replication dataErik Johnston2-13/+10
2017-04-04Add commentErik Johnston1-0/+5
2017-04-04Deduplicate new deviceinbox rows for replicationErik Johnston1-4/+7
2017-04-04Add the appropriate amount of preserve_fnErik Johnston4-17/+19
2017-04-04Fiddle tcp replication loggingErik Johnston1-2/+2
2017-04-04Advance replication streams even if nothing is listeningErik Johnston3-6/+15
Otherwise the streams don't advance and steadily fall behind, so when a worker does connect either a) they'll be streamed lots of old updates or b) the connection will fail as the streams are too far behind.
2017-04-04Remove unused worker config optionErik Johnston1-1/+0
2017-04-04Shuffle and comment synchrotron presenceErik Johnston1-9/+11
2017-04-04Move where we ack federationErik Johnston1-7/+16
2017-04-03Accept join events from all serversRichard van der Hoff2-5/+19
Make sure that we accept join events from any server, rather than just the origin server, to make the federation join dance work correctly. (Fixes #1893).
2017-04-03Remove spurious yieldRichard van der Hoff1-1/+1