-
-
Notifications
You must be signed in to change notification settings - Fork 677
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
History visibility database changes #2533
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Overall looks OK but some possible things we might want to look more at.
@@ -104,7 +104,7 @@ const selectEventsWithEventIDsSQL = "" + | |||
// the rowsToStreamEvents expects there to be exactly six columns. We need to | |||
// figure out if these really need to be in the DB, and if so, we need a | |||
// better permanent fix for this. - neilalexander, 2 Jan 2020 | |||
"SELECT event_id, added_at, headered_event_json, 0 AS session_id, false AS exclude_from_sync, '' AS transaction_id" + | |||
"SELECT event_id, added_at, headered_event_json, 0 AS session_id, false AS exclude_from_sync, '' AS transaction_id, 3 as history_visibility" + |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please comment what's happening here. I'm guessing it's to ensure all current state events are visible?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's because rowsToStreamEvents
is used in several places and, as the comment mentions, expects 7 columns and current_room_state
doesn't have a history_visibility
column.
We might need to query the actual history_visibility
from the output_room_events
table to get the correct value or would it be better to also add the column to the current_room_state
table (or a join on output_room_events
)?
@@ -101,9 +101,14 @@ func NewDatabase(base *base.BaseDendrite, dbProperties *config.DatabaseOptions) | |||
m := sqlutil.NewMigrations() | |||
deltas.LoadFixSequences(m) | |||
deltas.LoadRemoveSendToDeviceSentColumn(m) | |||
deltas.LoadAddHistoryVisibilityColumn(m) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Is this ordering (before the events table is created) right?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Oh, good catch!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hm.. I think the ordering is correct. We first create the table (separate db.Exec
statement at the top of NewDatabase
), execute the migrations and prepare the statements afterwards. If we move New*EventsTable
to be executed before the migrations, it will fail because history_visibilty
is missing (at least when coming from earlier versions).
syncapi/storage/shared/syncserver.go
Outdated
) (pduPosition types.StreamPosition, returnErr error) { | ||
returnErr = d.Writer.Do(d.DB, nil, func(txn *sql.Tx) error { | ||
var err error | ||
pos, err := d.OutputEvents.InsertEvent( | ||
ctx, txn, ev, addStateEventIDs, removeStateEventIDs, transactionID, excludeFromSync, | ||
ctx, txn, ev, addStateEventIDs, removeStateEventIDs, transactionID, excludeFromSync, historyVisibility.NumericValue(), |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
FWIW, we can implement sql.Scanner
and driver.Valuer
on the gomatrixserverlib.HistoryVisibility
type if we want to transform/parse the value at the database/sql
boundary, so that we don't need to have things like .NumericValue()
and .HistoryVisibilityFromInt()
and the intermediate uint8
s.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Done in matrix-org/gomatrixserverlib#326
evt *gomatrixserverlib.HeaderedEvent | ||
eventBytes []byte | ||
evt *gomatrixserverlib.HeaderedEvent | ||
historyVisibility uint8 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
... like here.
syncapi/storage/tables/interface.go
Outdated
addState, removeState []string, | ||
transactionID *api.TransactionID, | ||
excludeFromSync bool, | ||
historyVisibility uint8, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
... because then we can pass gomatrixserverlib.HistoryVisibility
types everywhere instead of having to type assert.
* Correctly redact events over federation (matrix-org#2526) * Ensure we check powerlevel/origin before redacting an event * Add passing test * Use pl.UserLevel * Make check more readable, also check for the sender * Add new next steps page to the documentation * Highlighting in docs * Rename the page to "Optimise your installation" * Attempt to raise the file descriptor limit at startup (matrix-org#2527) * Add `--difference` to `resolve-state` tool * Make the linter happy again * generic CaddyFile in front of Dendrite (monolith) (matrix-org#2531) for Caddy 2.5.x Co-authored-by: emanuele.aliberti <emanuele.aliberti@mtka.eu> * Handle state before, send history visibility in output (matrix-org#2532) * Check state before event * Tweaks * Refactor a bit, include in output events * Don't waste time if soft failed either * Tweak control flow, comments, use GMSL history visibility type * Fix rare panic when returning user devices over federation (matrix-org#2534) * Add `InputDeviceListUpdate` to the keyserver, remove old input API (matrix-org#2536) * Add `InputDeviceListUpdate` to the keyserver, remove old input API * Fix copyright * Log more information when a device list update fails * Fix nats.go commit (matrix-org#2540) Signed-off-by: Jean Lucas <jean@4ray.co> * Don't return `end` if there are not more messages (matrix-org#2542) * Be more spec compliant * Move lazyLoadMembers to own method * Return an error if trying to invite a malformed user ID (matrix-org#2543) * Add `evacuateUser` endpoint, use it when deactivating accounts (matrix-org#2545) * Add `evacuateUser` endpoint, use it when deactivating accounts * Populate the API * Clean up user devices when deactivating * Include invites, delete pushers * Silence presence logs (matrix-org#2547) * Blacklist `Guest users can join guest_access rooms` test until it can be investigated * Disable WebAssembly builds for now * Try to fix backfilling (matrix-org#2548) * Try to fix backfilling * Return start/end to not confuse clients * Update GMSL * Update GMSL * Roomserver producers package (matrix-org#2546) * Give the roomserver a producers package * Change init point * Populate ACLs API * Fix build issues * `RoomEventProducer` naming * Version 0.8.9 (matrix-org#2549) * Version 0.8.9 * Update changelog * feat+fix: Ignore unknown keys and verify required fields are present in appservice registration files (matrix-org#2550) * fix: ignore unknown keys in appservice configs fixes matrix-org#1567 * feat: verify required fields in appservice configs * Use new testrig for key changes tests (matrix-org#2552) * Use new testrig for tests * Log the error message * Fix QuerySharedUsers for the SyncAPI keychange consumer (matrix-org#2554) * Make more use of base.BaseDendrite * Fix QuerySharedUsers if no UserIDs are supplied * Return clearer error when no state NID exists for an event (matrix-org#2555) * Wrap error from `SnapshotNIDFromEventID` * Hopefully fix read receipts timestamps (matrix-org#2557) This should avoid coercions between signed and unsigned ints which might fix problems like `sql: converting argument $5 type: uint64 values with high bit set are not supported`. * Fix nil pointer access when redacting events (matrix-org#2560) * Fix issue `uint64 values with high bit are not supported` in presence (matrix-org#2562) * Fix issue matrix-org#2528 * Use gomatrixserverlib.Timestamp * Use ParseUint instead of ParseInt * Update Pinecone to matrix-org/pinecone@1ce778f * Ristretto cache (matrix-org#2563) * Try Ristretto cache * Tweak * It's beautiful * Update GMSL * More strict keyable interface * Fix that some more * Make less panicky * Don't enforce mutability checks for now * Determine mutability using deep equality * Tweaks * Namespace keys * Make federation caches mutable * Update cost estimation, add metric * Update GMSL * Estimate cost for metrics better * Reduce counters a bit * Try caching events * Some guards * Try again * Try this * Use separate caches for hopefully better hash distribution * Fix bug with admitting events into cache * Try to fix bugs * Check nil * Try that again * Preserve order jeezo this is messy * thanks VS Code for doing exactly the wrong thing * Try this again * Be more specific * aaaaargh * One more time * That might be better * Stronger sorting * Cache expiries, async publishing of EDUs * Put it back * Use a shared cache again * Cost estimation fixes * Update ristretto * Reduce counters a bit * Clean up a bit * Update GMSL * 1GB * Configurable cache sizees * Tweaks * Add `config.DataUnit` for specifying friendly cache sizes * Various tweaks * Update GMSL * Add back some lazy loading caching * Include key in cost * Include key in cost * Tweak max age handling, config key name * Only register prometheus metrics if requested * Review comments @S7evinK * Don't return errors when creating caches (it is better just to crash since otherwise we'll `nil`-pointer exception everywhere) * Review comments * Update sample configs * Update GHA Workflow * Update Complement images to Go 1.18 * Remove the cache test from the federation API as we no longer guarantee immediate cache admission * Don't check the caches in the renewal test * Possibly fix the upgrade tests * Update to matrix-org/gomatrixserverlib#322 * Update documentation to refer to Go 1.18 * Minor SendToDevice fix (matrix-org#2565) * Avoid unnecessary marshalling if sending to the local server * Fix ordering of ToDevice messages * Revive SendToDevice test * Use `/v3` to request media from remote servers (update to matrix-org/gomatrixserverlib#324) * Pointerise `types.RoomInfo` in the cache so we can update it in-place in the latest events updater * Add a Troubleshooting page * Update `sytest-whitelist` * Use sync API database in `filterSharedUsers` (matrix-org#2572) * Add function to the sync API storage package for filtering shared users * Use the database instead of asking the RS API * Fix unit tests * Fix map handling in `filterSharedUsers` * Update 1_createusers.md (matrix-org#2571) * Update 1_createusers.md Added description on how to create user accounts when running in docker. * Update 1_createusers.md Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com> * Fix connection_string format in dendrite-sample.polylith.yaml (matrix-org#2574) * History visibility database changes (matrix-org#2533) * Add new history_visibility column * Update SQL queries to include history_visibility * Store the history visibilty calculated by the roomserver * Update GMSL * Update migrations * Fix migration * Update GMSL * Fix `go.sum` * Update GMSL to use sql.Scanner & sql.Valuer * Re-order migration/table creation * Update gomatrixserverlib * Add history_visibility column to current_room_state * Fix migrations * Return error instead of Fatal log Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com> * Tweak cache counters (matrix-org#2575) * Tweak cache counters This makes the number of counters relative to the maximum cache size. Since the counters effectively manage the size of the bloom filter, larger caches need more counters and smaller caches need less. 10 counters per 1KB data means that the default cache size of 1GB should result in a bloom filter and TinyLRU admission set of about 16MB estimated. * Remove line left by accident * Set historyVisibility in rowsToStreamEvents * Update FAQ * Add event state key cache (matrix-org#2576) * Explain how SRV works in Matrix and discourage using it (matrix-org#2577) * Explain how SRV works in Matrix and discourage using it * Minor tweaks to formatting Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com> * Fix issue with membership event_nid being 0 (matrix-org#2580) * docs: Add build page; correct proxy info; fix Caddy example (matrix-org#2579) * Add build page; correct proxy info; fix Caddy example * Improve Caddyfile example * Apply review comments; add polylith Caddyfile * Bump tzinfo from 1.2.9 to 1.2.10 in /docs (matrix-org#2584) Bumps [tzinfo](https://github.com/tzinfo/tzinfo) from 1.2.9 to 1.2.10. - [Release notes](https://github.com/tzinfo/tzinfo/releases) - [Changelog](https://github.com/tzinfo/tzinfo/blob/master/CHANGES.md) - [Commits](tzinfo/tzinfo@v1.2.9...v1.2.10) --- updated-dependencies: - dependency-name: tzinfo dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * Membership updater refactoring (matrix-org#2541) * Membership updater refactoring * Pass in membership state * Use membership check rather than referring to state directly * Delete irrelevant membership states * We don't need the leave event after all * Tweaks * Put a log entry in that I might stand a chance of finding * Be less panicky * Tweak invite handling * Don't freak if we can't find the event NID * Use event NID from `types.Event` * Clean up * Better invite handling * Placate the almighty linter * Blacklist a Sytest which is otherwise fine under Complement for reasons I don't understand * Fix the sytest after all (thanks @S7evinK for the spot) * Try to fix HTTP 500s on `/members` (matrix-org#2581) * Update database migrations, remove goose (matrix-org#2264) * Add new db migration * Update migrations Remove goose * Add possibility to test direct upgrades * Try to fix WASM test * Add checks for specific migrations * Remove AddMigration Use WithTransaction Add Dendrite version to table * Fix linter issues * Update tests * Update comments, outdent if * Namespace migrations * Add direct upgrade tests, skipping over one version * Split migrations * Update go version in CI * Fix copy&paste mistake * Use contexts in migrations Co-authored-by: kegsay <kegan@matrix.org> Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com> * Add .well-known/matrix/client to clientapi (matrix-org#2551) Signed-off-by: Jonathan Bartlett <jonathan@jonnobrow.co.uk> Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com> * Remove `room_id` field from MSC2946 stripped events (closes matrix-org#2588) * Remove `goose` from Dockerfiles * Make the User API responsible for sending account data output events (matrix-org#2592) * Make the User API responsible for sending account data output events * Clean up producer * Review comments * Update NATS Server and nats.go to use upstream * Set CORS headers for HTTP 404 and 405 errors (matrix-org#2599) * Set CORS headers for the 404s * Use custom handlers, plus one for HTTP 405 too * Tweak setup * Add to muxes too * Tidy up some more * Use built-in HTTP 404 handler * Don't bother setting it for federation-facing * Optimise checking other servers allowed to see events (matrix-org#2596) * Try optimising checking if server is allowed to see event * Fix error * Handle case where snapshot NID is 0 * Fix query * Update SQL * Clean up `CheckServerAllowedToSeeEvent` * Not supported on SQLite * Maybe placate the unit tests * Review comments * De-race `types.RoomInfo` (matrix-org#2600) * De-race `CompleteSync` (matrix-org#2601) The `err` was coming from outside of the goroutine and being written to by concurrent goroutines. * Version 0.9.0 (matrix-org#2602) Co-authored-by: Till <2353100+S7evinK@users.noreply.github.com> Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com> Co-authored-by: Till Faelligen <davidf@element.io> Co-authored-by: Emanuele Aliberti <dev@mtka.eu> Co-authored-by: emanuele.aliberti <emanuele.aliberti@mtka.eu> Co-authored-by: Jean Lucas <jean@4ray.co> Co-authored-by: Kabir Kwatra <kabir@kwatra.me> Co-authored-by: andreever <52261463+andreever@users.noreply.github.com> Co-authored-by: Maximilian Gaedig <38767445+MaximilianGaedig@users.noreply.github.com> Co-authored-by: Tulir Asokan <tulir@maunium.net> Co-authored-by: Matt Holt <mholt@users.noreply.github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: kegsay <kegan@matrix.org> Co-authored-by: Jonathan Bartlett <34320158+Jonnobrow@users.noreply.github.com>
Needs matrix-org/gomatrixserverlib#318
As per the spec, "shared" is the default value, if it is unknown. To avoid opening up conversations we already stored, the migrations use "joined" for
m.room.message
andm.room.encrypted
instead.