`AppState.hs` is imported by many modules. Some of those modules contain
important functions that need to be imported back to `AppState.hs`, causing
circular dependency problem.
Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
This gives the import statements more space, essentially, it collectively
removes 90+ unneeded lines in the imports statements.
Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
src/ now contains all source code - in subdirectories, according to the
.cabal component they belong to. This will allow us to put vendored
libraries in the same place - and later split our own code into multiple
components/libraries as well.
Runs the doctests much faster, which is potentially useful in
combination with postgrest-watch for local development.
This implies that doctests run on compiled code, not in a GHCi session,
which has some implications:
- Only exported functions can be tested.
- Imports need to be made explicit in doctests themselves.
On the flipside, this would allow us to potentially include doctest
results in code coverage, I believe.
This change is a requirement to vendor hasql, which otherwise breaks the
existing doctests: hasql contains a .hsc file, which *needs* to be
compiled - not interpreted - to make the tests work.
Adds the `url_use_legacy_target_names` config.
Enabled (default):
* It allows using the resource name in filters,
orders or limits when it has an alias, e.g.
`table?select=alias:target(*)&target.id=eq.1`
* Logs a WARNING with a hint to use the alias
* Returns a Warning header in the response
Disabled:
* It returns an error, only the alias is allowed
* No warnings returned
This feature is deprecated
This change introduces a new way to perform liveness check - instead of trying to open a TCP connection to main server, we check if main socket is listening and if main server accept loop thread is alive.
Opening a TCP connection in liveness check was problematic because:
* it used available file descriptors which might have been a problem under load
* made liveness check unreliable when multiple PostgREST instances are available on the same port (eg. using SO_REUSEPORT)
Right now code in App.run does not properly use bracket/finally to close sockets and clean-up mainSocketRef. This is not a big problem at the moment because the application is going to exit enyway but introducting proper resource handling will make future refactorings safer.
This change makes AppState and Listener modules independent from the way how application is terminated. It removes stateMainThreadId from AppState and introduces stateKillApp. It also removes exported function AppState.getMainThreadId replacing it with more general killApp.
This change makes Admin module independent from how liveness check is performed. It moves reachMainApp function to App module and makes Admin.runAdmin, instead of action to get the main socket, take an IO action checking liveness as a parameter.
Thanks to this, any change in how liveness check is performed and, more importantly, what data it uses, does not require changing multiple modules (Admin and App).
Admin server was crashing silently and requests hang indefinitely.
With this fix, we are now logging the admin server exceptions and also
close the socket afterwards so admin request don't hang.
Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
BREAKING CHANGE
Breaks the string comparison operators implemented in #3813. Those can
be replaced with regex searches using JSON Path `search()` function.
Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
For admin server, port server-setting was set to the same port as
app server port. This doesn't have any effect because it is overridden
by the already initialized admin server socket, but it's better to
explicitly correct it to avoid anything unexpected.
Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
This change ensures PostgREST starts listening on a server socket only after it loaded the schema cache and is ready to handle requests. It is no longer going to return 503 errors during startup until the schema cache is loaded.
Right now stateNextListenerDelay is only used in Listener module.
This change is a small refactoring moving nextListenerDelay to Listener to increase cohesion and decrease coupling.
This is useful when multiple instances run on the same machine, for
example behind a proxy. Unix sockets for web and admin servers can then
be put in the same folder for each instance.
Can be helpful when writing tests as well.
retryingSchemaCacheLoad should not clear existing schema cache upon failure - there is no reason to do that. If there is a communication issue with the database server or db is down, clients are going to get 502 anyway. If it was a glitch when loading the schema cache - the clients are going to use old (stale) schema cache for some time until next retry re-loads it successfully.
Replaces the `Maybe [Text]` with `[Text]`. The `Maybe` is unnecessary
because we handle `Just []` and `Nothing` the same way.
Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
The PGRST000 database connection error message was "Database connection
error. Retrying the connection.", but reconnection attempts are already
logged separately by the reconnection observation, and on fatal errors
(e.g. authentication failure) PostgREST does not retry at all. Drop the
"Retrying the connection." part, leaving "Database connection error.".
The config dump in Config.hs listed db-config, db-pre-config and the
db-pool-* settings out of alphabetical order, while the rest of the dump
was sorted. Now the whole dump is consistently alphabetical and update
the expected IO test configs to match.
Currently, authentication and response execution each unwrap ExceptT with separate runExceptT calls, which split the main request flow across nested pattern matching and Either handling. Control flow is complex and difficult to understand.
The goal of this change is to make request execution as sequential
monadic code with clear error handling.
To implement that, request handling is now run in ExceptT over WriterT (Last ByteString) IO monad stack. Auth role is written after authentication succeeds and further returned along the response. Thanks to it response observation generation is centralized at the end of request handling.
It was necessary to abstract monad stack in getAuthResult, lookupJwtCache, postgrestResponse, and withTiming to enable introduction of WriterT.
This change makes the API surface between MainTx and App smaller.
Currently, App reconstructs a database transaction by unpacking the isolation
level, transaction mode, DbHandler, and transaction runner returned by MainTx.
That exposes MainTx internals at the call site even though MainTx already owns
query setup, execution, decoding, and rollback behavior.
The goal is to keep transaction assembly in MainTx while App remains responsible
for pool execution, database error mapping, and response orchestration. DbTx now
carries the assembled SQL session, and App passes that session directly to the
connection pool.
Schema cache query timings are only needed immediately after a schema-cache reload to emit SchemaCacheQueriedObs. Storing them inside SchemaCache makes the cache carry transient observability data that is not part of the cached schema state and is never used by request handling.
This change makes querySchemaCache to return query timings in a tuple in parallel to SchemaCache and removes dbQueryTimings field.
This change gets rid of unnecessary explicit bindRandomPortTCP in initServerSocket. Returned port value was ignored in removed code anyway as assigned port retrieval from an open socket is handled elsewhere.
Remove internal schema cache load and relationship load sleep settings plus
the delay wrappers they enabled. Drop IO tests that depended on the removed
settings.
We decided against doing this in #4913, therefore removing it. With that,
also mentioning that control flow never reaches there to avoid confusion.
Signed-off-by: Taimoor Zaeem <taimoorzaeem@gmail.com>
Right now metrics observation handler does not track database connections but updates a single Gauge based on HasqlPoolObs events. This is problematic because Hasql pool reports various connection events in multiple phases. The connection state machine is not simple and to precisely report the number of connections in various states, it is necessary to track their lifecycles.
This change adds a ConnTrack data structure and logic to track database connections lifecycles. At the moment it supports "connected" and "inUse" connection counts precisely. The "pgrst_db_pool_available" metric is implemented on top of ConnTrack instead of a simple Gauge.
Disabling the autovacuum daemon should also help reproducibility in
theory, although I don't know of any cases where we hit a problem with
that.
VACUUM changes the order of rows that PostgreSQL returns for some table
without explicit ordering, thus doing the latter to make it consistently
reproducible.
After ANALYZE estimates are 100% exact for the moment, so some requests
which returned 206 Partial Response now return 200 instead. The fact
that PostgREST returns 206 on an unfiltered endpoint can probably be
considered a bug.
Taimoor Zaeem