Summary:
Able to run 1000 tasks with object dependencies on a set of distributed Raylets.
Raylet Changes:
Finalized ClientConnection class.
Task forwarding.
NM-to-NM heartbeats.
NM resource accounting for tasks.
Simple scheduling policy with task forwarding.
Creating and maintaining NM 2 NM long-lived connections and reusing them for task forwarding.
LineageCache Changes:
LineageCache without cleanup of tasks committed by remote nodes.
Lineage cache writeback and cleanup implementation.
ObjectManager Changes:
Object manager event loop/ClientConnection refactor.
Multithreaded object manager (disabled in this PR).
Testing Changes:
Integration tests for task submission on multiple Raylets.
Stress tests for object manager (with GCS and object store integration).
Co-authored-by: Stephanie Wang <swang@cs.berkeley.edu>
Co-authored-by: Alexey Tumanov <atumanov@gmail.com>
* TABLE_APPEND call
* Convert callbacks back to taking in a string...
* GCS returns flatbuffers, define Log class
* Cleanups
* Modify client table to use the Log interface
* Fix bug where we replied twice from redis
* Fixes
* lint
* Compile and test raylet TaskTable
* Modify GCS tables to handle unique_ptrs from nested flatbuffers
* Add raylet::TaskTable unit tests to replace ObjectTable ones
* Convert ObjectTable to a log
* Convert ObjectTable tests to the Log
* AppendAt Redis and gcs Log command
* unit test for AppendAt
* Add a Log for task reconstruction data
* Add check for unique entries in TABLE_APPEND
* Documentation
* TABLE_APPEND call
* Convert callbacks back to taking in a string...
* GCS returns flatbuffers, define Log class
* Cleanups
* Modify client table to use the Log interface
* Fix bug where we replied twice from redis
* Fixes
* lint
* Add TableRequestNotifications and TableCancelNotifications to Redis modules
* Add RequestNotifications and CancelNotifications to generic GCS Table
* Add tests for subscribing to specific keys
* Remove TODO!
* Return the current value at the key directly from RequestNotifications instead of through publish
* Add unit test for Lookup failure callback
* Modify tests to account for empty subscription response
* Remove ObjectTable notification methods
* Clean up message parsing and doc in redis context
* Use vectors of DataT in all GCS callbacks
* Clean up SubscriptionCallback
* Move Table definitions into tables.cc
* Refactor and document redis modules
* doc
* Fix new GCS build
* Cleanups
* Revert "Fix new GCS build"
This reverts commit 6e3e69090c67ef60aaf22a9cf62be0290d989e96.
* Use vectors for internal callback interface, user-facing interface takes a reference to a single item
* Fix new GCS build
* Add unit test for Lookup failure callback
* Fix compiler errors
* Cleanup
* Publish the entry ID with the notification
* Check that the ID for a notification matches in client tests
* Define execution dependencies flatbuffer and add to Redis commands
* Convert TaskSpec to TaskExecutionSpec
* Add execution dependencies to Python bindings
* Submitting actor tasks uses execution dependency API instead of dummy argument
* Fix dependency getters and some cleanup for fetching missing dependencies
* C++ convention
* Make TaskExecutionSpec a C++ class
* Convert local scheduler to use TaskExecutionSpec class
* Convert some pointers to references
* Finish conversion to TaskExecutionSpec class
* fix
* Fix
* Fix memory errors?
* Cast flatbuffers GetSize to size_t
* Fixes
* add more retries in global scheduler unit test
* fix linting and cast fbb.GetSize to size_t
* Style and doc
* Fix linting and simplify from_flatbuf.
* Object table lookup returns vector of DBClientID instead of address strings
* Add node IP address to DBClient notification
* DB client cache stores entire DB client, convert addresses to std::string
* get cached db client returns the client
* Expose a call to initialize the redis cache
* Local scheduler filters out dead clients during reconstruction
* Remove node ip address from dbclient, use aux_address for plasma managers
* Get entire db client entry when not found in cache
* Fix common tests
* Fix address in tests
* Push error to driver if driver task did the put
* Address Robert's comments and cleanup
* Remove unused Redis command
* Fix db test
* WIP: removing OL, OI, TT on client exit; no saving yet.
* ray_redis_module.cc: update header comment.
* Cleanup: just the removal.
* Reformat via yapf: use pep8 style instead of google.
* Checkpoint addressing comments (partially)
* Add 'b' marker before strings (py3 compat)
* Add MonitorTest.
* Use `isort` to sort imports.
* Remove some loggings
* Fix flake8 noqa marker runtest.py
* Try to separate tests out to monitor_test.py
* Rework cleanup algorithm: correct logic
* Extend tests to cover multi-shard cases
* Add some small comments and formatting changes.
* Remove race between local scheduler disconnecting and global scheduler
assigning a task
* Fix number of workers started in component failures test
* Fix race between global scheduler retrying a task assignment and monitor
cleaning up task table. The global scheduler should only retry the task
assignment if the local scheduler is still alive.
* Clean up task_table_update callback if failure
* Look up current local scheduler mapping when retrying actor task submission
* Log warning if no subscribers received a task table update
* Clean up database handle memory in local scheduler
* Replace a local scheduler ut_array with a std::vector.
* Replace vector of sizes in local scheduler with std::pair.
* Remove utarray include.
* Replace utarray with std::vector for reading local scheduler input messages.
* Remove more UT data structures.
* Remove UT includes.
* Fix linting.
* Include stdlib.h to find size_t.
* Remove includes of stdbool.h.
* Replace std::pair with TaskQueueEntry.
* Fix redis tests.
* Reinstate tests.
* added log_table function and a test
* fixed log_files and added task_profiles
* fixed formatting
* fixed linting errors
* fixes
* removed file
* more fixes
* hopefully fixed
* Small changes.
* Fix linting.
* Fix bug in log monitor.
* Small changes.
* Fix bug in travis.
* Including data_size and hash in the ResultTableReply.
* Included data_size and hash info in object_table.
* Fixed bugs in ray_redis_module.cc.
* Removing commented out code.
* Fixes
* Freed hash and data_size strings after using, and checked if they're null along with task_id and is_put.
* Changed it so that data_size is set correctly.
* Removed iostream import.
* Included a check to ensure that the Redis string to long long conversion was successful.
* Included separate data_size and hash null checks.
* Fixed bug.
* Made linting changes.
* Another linting error.
* Slight simplication.
* Implement sharding in the Ray core
* Single node Python modifications to do sharding
* Do the sharding in redis.cc
* Pipe num_redis_shards through start_ray.py and worker.py.
* Use multiple redis shards in multinode tests.
* first steps for sharding ray.global_state
* Fix problem in multinode docker test.
* fix runtest.py
* fix some tests
* fix redis shard startup
* fix redis sharding
* fix
* fix bug introduced by the map-iterator being consumed
* fix sharding bug
* shard event table
* update number of Redis clients to be 64K
* Fix object table tests by flushing shards in between unit tests
* Fix local scheduler tests
* Documentation
* Register shard locations in the primary shard
* Add plasma unit tests back to build
* lint
* lint and fix build
* Fix
* Address Robert's comments
* Refactor start_ray_processes to start Redis shard
* lint
* Fix global scheduler python tests
* Fix redis module test
* Fix plasma test
* Fix component failure test
* Fix local scheduler test
* Fix runtest.py
* Fix global scheduler test for python3
* Fix task_table_test_and_update bug, from actor task table submission race
* Fix jenkins tests.
* Retry Redis shard connections
* Fix test cases
* Convert database clients to DBClient struct
* Fix race condition when subscribing to db client table
* Remove unused lines, add APITest for sharded Ray
* Fix
* Fix memory leak
* Suppress ReconstructionTests output
* Suppress output for APITestSharded
* Reissue task table add/update commands if initial command does not publish to any subscribers.
* fix
* Fix linting.
* fix tests
* fix linting
* fix python test
* fix linting
* Fix worker blocked bug
* tmp
* Push an error to the driver on ray.put for non-driver tasks
* Fix result table tests
* Fix test, logging
* Address comments
* Fix suppression bug
* Fix redis module test
* Edit error message
* Get values in chunks during reconstruction
* Test case for driver ray.put errors
* Error for evicting ray.put objects from the driver
* Fix tests
* Reduce verbosity
* Documentation
* Failing test case
* Local scheduler exits cleanly after plasma store dies
* Tolerate one plasma store failure
* Tolerate plasma store failures on all nodes except head node
* Plasma manager heartbeats
* Component failure tests
* Don't run the helper for Python testing
* Fix C test
* Fix hanging plasma transfer test
* Fix python3
* Consolidate ClientConnection code
* Fix valgrind test
* fix c test
* We can restart worker nodes!
* Fix flatbuffers bug
* Address comments
* Only register actual workers with the local scheduler
* Fix bug
* Fix segfaults
* Add test case that tests for driver liveness, fix local scheduler bug
* Clean up after tests
* Allocate retry info on the stack
* Send SIGKILL before waiting
* Relax unit test conditions
* Driver liveness test case and documentation
* Compile the Ray redis module with C++.
* Redo parsing of object table notifications with flatbuffers.
* Update redis module python tests.
* Redo parsing of task table notifications with flatbuffers.
* Fix linting.
* Redo parsing of db client notifications with flatbuffers.
* Redo publishing of local scheduler heartbeats with flatbuffers.
* Fix linting.
* Remove usage of fixed-width formatting of scheduling state in channel name.
* Reply with flatbuffer object to task table queries, also simplify redis string to flatbuffer string conversion.
* Fix linting and tests.
* fix
* cleanup
* simplify logic in ReplyWithTask
* WARN instead of FATAL for object hash mismatches, push error to driver
* Document the callback signature for object_table_add/remove
* Error table
* Wait for all errors in python test
* Fix doc
* Fix state test
* Clean up plasma subscribers on EPIPE
First pass at a monitoring script - monitor can detect local scheduler death
Clean up task table upon local scheduler death in monitoring script
Don't schedule to dead local schedulers in global scheduler
Have global scheduler update the db clients table, monitor script cleans up state
Documentation
Monitor script should scan tables before beginning to read from subscription channel
Fix for python3
Redirect monitor output to redis logs, fix hanging in multinode tests
* Publish auxiliary addresses as part of db_client deletion notifications
* Fix test case?
* Small changes.
* Use SCAN instead of KEYS
* Address comments
* Address more comments
* Free redis module strings
* Availability after a killed worker
* Workers exit cleanly
* Memory cleanup in photon C tests
* Worker failure in multinode
* Consolidate worker cleanup handlers
* Update the result table before handling a task submission
* KILL_WORKER_TIMEOUT -> KILL_WORKER_TIMEOUT_MILLISECONDS
* Log a warning instead of crashing if no result table entry found
* First pass at reconstruction in the worker
Modify reconstruction stress testing to start Plasma service before rest of Ray cluster
TODO about reconstructing ray.puts
Fix ray.put error for double creates
Distinguish between empty entry and no entry in object table
Fix test case
Fix Python test
Fix tests
* Only call reconstruct on objects we have not yet received
* Address review comments
* Fix reconstruction for Python3
* remove unused code
* Address Robert's comments, stress tests are crashing
* Test and update the task's scheduling state to suppress duplicate
reconstruction requests.
* Split result table into two lookups, one for task ID and the other as a
test-and-set for the task state
* Fix object table tests
* Fix redis module result_table_lookup test case
* Multinode reconstruction tests
* Fix python3 test case
* rename
* Use new start_redis
* Remove unused code
* lint
* indent
* Address Robert's comments
* Use start_redis from ray.services in state table tests
* Remove unnecessary memset
* Switch to using redis modules for task table.
* Switch to using redis modules for the task table.
* Fix some tests.
* Fix naming and remove code duplication.
* Remove duplication in redis modules and add more cleanups.
* Address comments.
* Add function for driver to get address info from Redis.
* Use Redis address instead of Redis port.
* Configure Redis to run in unprotected mode.
* Add method for starting Ray processes on non-head node.
* Pass in correct node ip address to start_plasma_manager.
* Script for starting Ray processes.
* Handle the case where an object already exists in the store. Maybe this should also compare the object hashes.
* Have driver get info from Redis when start_ray_local=False.
* Fix.
* Script for killing ray processes.
* Catch some errors when the main_loop in a worker throws an exception.
* Allow redirecting stdout and stderr to /dev/null.
* Wrap start_ray.py in a shell script.
* More helpful error messages.
* Fixes.
* Wait for redis server to start up before configuring it.
* Allow seeding of deterministic object ID generation.
* Small change.
* adding object broadcast channel; published on each object table add
* publishing data size to the bcast channel
* bug fix: objectkey
* update object tests to test for data size: C + py
* remove debug
* clang format
* Minor changes.
* Fix error.
* merging with Robert's comments
* clang format for the object table test upgrade
* Task table redis module implementation
* Publish tasks and take in individual fields as args, not task object
* Scheduling state integer has width 1, error on illegal put
* Unit tests for task table and more documentation
* Task table subscribe, fix publish topics and address Philipp and Alexey's comments
* Helper function to create prefixed strings
* Factor out the table prefixes in the test cases
* Add RAY.CONNECT Redis command.
* Add RAY.GET_CLIENT_ADDRESS command.
* Build and clean Redis in common Makefile.
* Use custom Redis module in Ray and use custom CONNECT and GET_CLIENT_ADDRESS commands.
* Fixes.
* Remove mapping from redis client ID to ray db client ID.
* Fix.
* initial redis module
* temp commit
* temp commit
* temp commit
* Empty object table functions and broken object_table_lookup
* fix segfault and clean up code
* cleanup and tests
* try to ignore redismodule.h
* check if data_size is integer
* Minor changes to redis-module tests.
* try to exclude redismodule from clang-format
* try something different
* fix clang-format and tests
* sleep a bit
* Result table
* fix redis_module tests
* fix tests and add tests for result table
* more tests
* randomize ports
* Minor changes.
* More fixes.
