* test: improves testing harness for engine
* update CI test
* fix: race condition in test
* make tests more stable
* cleanup pub and sub buffers
* fix: goleak on rampup test
* feat: matrix tests for engine
* clean up rabbit mq session stuff, add a quick ack and error processing for AddMessage
* bit more paranoid about getting stuck in chans
* first pass at locking the message to deal with the failed states better
* clean up the access to ready for the mq
* make sure we don't block sending this ack
- Simplifies architecture for splitting engine services into different components. The three supported services are now `grpc-api`, `scheduler`, and `controllers`. The `grpc-api` service is the only one which needs to be exposed for workers. The other two can run as unexposed services.
- Fixes a set of bugs and race conditions in the `v2` scheduler
- Adds a `lastActive` time to the `Queue` table and includes a migration which sets this `lastActive` time for the most recent 24 hours of queues. Effectively this means that the max scheduling time in a queue is 24 hours.
- Rewrites the `ListWorkflowsForEvent` query to improve performance and select far fewer rows.
Refactors the queueing logic to be fairly balanced between actions, with each action backed as a separate FIFO queue. Also adds support for priority queueing and custom queues, though those aren't exposed on the API layer yet. Improves throughput to be > 5000 tasks/second on a single queue.
---------
Co-authored-by: Alexander Belanger <alexander@hatchet.run>
* feat: allow extending the api server
* chore: remove internal packages to pkg
* chore: update db_gen.go
* fix: expose auth
* fix: move logger to pkg
* fix: don't generate gitignore for prisma client
* fix: allow extensions to register their own api spec
* feat: expose pool on server config
* fix: nil pointer exception on empty opts
* fix: run.go file
* feat: add rabbitmq connection pool and remove non-fatal worker errors
* chore: go mod tidy
* fix: release pool after opening channel
* fix: make sure channel is closed after all tasks return on subscribe
* fix: don't loop endlessly
Logic for requeueing and reassigning did not limit the number of step runs to requeue, so when events accumulate with no worker present it causes memory to spike along with a very high query latency on the database. This commit limits the number of step runs returned in the requeue and reassign queries, and also properly locks step run rows for these queries so only a step run in a PENDING or PENDING_ASSIGNMENT state can be requeued.
It also improves performance of the `AssignStepRunToWorker` query and ensures that `maxRuns` on workers are always respected through the introduction of a `WorkerSemaphore` model. The value gets decremented when a step run is assigned and incremented when a step run is in a final state.
Co-authored-by: Luca Steeb <contact@luca-steeb.com>
* Update controller.go
---------
Co-authored-by: steebchen <contact@luca-steeb.com>
