Commit graph

2593 commits

Author SHA1 Message Date
Daniel Edgecumbe
a70ae1687b .gitignore: Add Vim swap files (#4016) 2019-02-11 10:27:10 -08:00
Ion
3c32343c63 Ray signal (#3624) 2019-02-11 10:14:48 -08:00
ebrevdo
52dfde1cbb Update flatbuffer bazel rule to work with flatbuffer master branch. (#4008) 2019-02-11 10:00:06 -08:00
Zhijun Fu
7097ba393b protect raylet against bad messages (#4003)
* protect raylet against bad messages

* address comments

* linting and regression test
2019-02-12 00:39:38 +08:00
Wang Qing
bc438ca73b [Java] Refine Java config item (#4014)
* Refine

* Address comment.
2019-02-11 23:55:40 +08:00
Philipp Moritz
ab809bd927 update ray version to 0.7.0dev (#3995) 2019-02-10 19:56:42 -08:00
Eric Liang
8e9f2c923f
[autoscaler] Use RLock in addition to FileLock 2019-02-10 19:16:43 -08:00
Yuhong Guo
5fb1efd60d Fix CI test failures (#4007) 2019-02-11 11:01:14 +08:00
bjg2
e703b9f49d [wingman -> rllib] Improved stats changes in AsyncSamplesOptimizer (#3966)
* added stats changes to optimizer

* changes timers

* fix python 2 compat

* improved optimizer throughput stats

* Update async_samples_optimizer.py

* fix python2 compat
2019-02-10 01:25:22 -08:00
Yuhong Guo
3a66d47a3a
Remove RAY_CHECK from JNI code (#3978)
* Remove RAY_CHECK in JNI

* Try to add mvn test to test the exception.

* Refine

* Address comments
2019-02-09 18:10:22 +08:00
bibabolynn
728031a972 [java] when put an object in plasma store, ignore "object alreay exists" exception (#3687)
* distinct plasma client exception

* Update ObjectStoreProxy.java

* Update and rename PlasmaArrowTest.java to PlasmaStoreTest.java

* store put

* Use testng to replace junit to fix test failure
2019-02-09 18:03:17 +08:00
Eric Liang
29322c7389
[rllib] Replay buffer for IMPALA should default to 0 slots. (#3971)
* disable replay

* make lq configurable

* leak test

* Update run_multi_node_tests.sh
2019-02-08 10:03:11 -08:00
Robert Nishihara
6a32b410bb Update versions from 0.6.2 -> 0.6.3 in the documentation. (#3981) 2019-02-07 20:57:37 -08:00
Robert Nishihara
ef527f84ab Stream logs to driver by default. (#3892)
* Stream logs to driver by default.

* Fix from rebase

* Redirect raylet output independently of worker output.

* Fix.

* Create redis client with services.create_redis_client.

* Suppress Redis connection error at exit.

* Remove thread_safe_client from redis.

* Shutdown driver threads in ray.shutdown().

* Add warning for too many log messages.

* Only stop threads if worker is connected.

* Only stop threads if they exist.

* Remove unnecessary try/excepts.

* Fix

* Only add new logging handler once.

* Increase timeout.

* Fix tempfile test.

* Fix logging in cluster_utils.

* Revert "Increase timeout."

This reverts commit b3846b89040bcd8e583b2e18cb513cb040e71d95.

* Retry longer when connecting to plasma store from node manager and object manager.

* Close pubsub channels to avoid leaking file descriptors.

* Limit log monitor open files to 200.

* Increase plasma connect retries.

* Add comment.
2019-02-07 19:53:50 -08:00
Philipp Moritz
0aa74fb1fd Update cloudpickle to 0.8.0.dev0 (#3964) 2019-02-07 15:24:06 -08:00
Eric Liang
ae4bc7d6e8
[revert] [rllib] Add copy() in async samples optimizer 2019-02-07 14:14:39 -08:00
markgoodhead
5ce670cb36 [tune] Add Initial Parameter Suggestion for HyperOpt (#3944)
Allows users of the HyperOptSearch suggestion algorithm to specify initial experiment values to run (typically already known good baseline parameters within the domain specified)
2019-02-07 10:57:51 -08:00
Ion
f987572795 Inline objects (#3756)
* added store_client_ to object_manager and node_manager

* half through...

* all code in, and compiling! Nothing tested though...

* something is working ;-)

* added a few more comments

* now, add only one entry to the in GCS for inlined objects

* more comments

* remove a spurious todo

* some comment updates

* add test

* added support for meta data for inline objects

* avoid some copies

* Initialize plasma client in tests

* Better comments. Enable configuring nline_object_max_size_bytes.

* Update src/ray/object_manager/object_manager.cc

Co-Authored-By: istoica <istoica@cs.berkeley.edu>

* Update src/ray/raylet/node_manager.cc

Co-Authored-By: istoica <istoica@cs.berkeley.edu>

* Update src/ray/raylet/node_manager.cc

Co-Authored-By: istoica <istoica@cs.berkeley.edu>

* fiexed comments

* fixed various typos in comments

* updated comments in object_manager.h and object_manager.cc

* addressed all comments...hopefully ;-)

* Only add eviction entries for objects that are not inlined

* fixed a bunch of comments

* Fix test

* Fix object transfer dump test

* lint

* Comments

* Fix test?

* Fix test?

* lint

* fix build

* Fix build

* lint

* Use const ref

* Fixes, don't let object manager hang

* Increase object transfer retry time for travis?

* Fix test

* Fix test?

* Add internal config to java, fix PlasmaFreeTest
2019-02-07 10:32:39 -08:00
Richard Liaw
5db1afef07
[tune] Support Custom Resources (#2979)
Support arbitrary resource declarations in Tune.

Fixes https://github.com/ray-project/ray/issues/2875
2019-02-07 00:29:19 -08:00
Robert Nishihara
a654152f9c Pin gym version in Python 2 tests. (#3973) 2019-02-06 23:56:14 -08:00
Philipp Moritz
3bb65677dc Use one memory mapped file for plasma (#3871) 2019-02-06 23:53:05 -08:00
Stephanie Wang
d2b6db3db1
Bump version from 0.6.2 to 0.6.3 (#3972) 2019-02-06 19:11:16 -08:00
Eric Liang
04fc145a44 [autoscaler] Autoscaler hangs forever on non-zero exit code command (#3969) 2019-02-06 17:25:24 -08:00
Stephanie Wang
49e9bec988
Fix raylet bug in driver cleanup (#3962)
* Fix task dependency manager cleanup on driver exit

* Add regression test

* Better check, update header
2019-02-06 11:19:10 -08:00
Stephanie Wang
244fd473f4
Only mark tasks as forwarded if they are in the lineage cache (#3958) 2019-02-05 23:01:38 -08:00
Alex LaGrassa
b0fe5af7c8 [doc] Update example-parameter-server.rst (#3773) 2019-02-05 22:00:54 -08:00
Robert Nishihara
fa4eb8313d Suppress warning for serializing different unique ID types in Python. (#3872)
* Suppress warning for serializing different unique ID types in Python.

* Add _ID_TYPES variable.
2019-02-05 11:38:33 -08:00
vfdev
b2b8417790 [tune] Improve mnist_pytorch.py example (#3894)
## What do these changes do?

* Improved --no-cuda handling
* Removed deprecated Variable usage


## Related issue number

Fixes #3873 
<!-- Are there any issues opened that will be resolved by merging this change? -->
2019-02-04 17:59:54 -08:00
Eric Liang
5fb813ff39
Don't check fail on missing lineage cache entry (#3861) 2019-02-04 17:45:41 -08:00
William Ma
f067223c4a Allow Ray processes to be started inside of gdb and tmux. (#3847) 2019-02-04 15:23:39 -08:00
Yuhong Guo
add8ae7063 Add bazel build for JNI code (#3918)
* Add bazel build for JNI code

* clean

* Add plasma client JNI build process

* refine

* clean linux part

* Add Java Library

* Remove java library

* Generate dylib after build using genrule
2019-02-04 13:03:46 -08:00
Wang Qing
e1c68a0881 Enable including Java worker for ray start command (#3838) 2019-02-04 16:23:43 +08:00
Eric Liang
7ef830bef1 [rllib] Add copy() in async samples optimizer to fix memory leak (#3938)
Fixes #3884.
2019-02-03 18:34:37 -08:00
Andrew Tan
8323419a6d [tune] Add SigOpt Integration (#3844) 2019-02-03 18:23:57 -08:00
Kristian Hartikainen
85294fb503 [autoscaler] node caching changes (#3937)
Breaks the node provider node getter into cached and non-cached versions.

Fixes #3930 by updating the node label finger print before updating labels.
Fixes #3935 by refreshing node cache if node ip is not found.
2019-02-03 17:48:07 -08:00
James Casbon
976f018dab [autoscaler] GCP: only call setIamPolicy if necessary (#3782) 2019-02-03 16:16:00 -08:00
James Casbon
b8cc176b4d [autoscaler] Document gcp subnet config (#3783)
Adds info to the gcp example yaml on using shared subnets.
2019-02-03 16:14:44 -08:00
Si-Yuan
9295ab8f60 Various Python code cleanups. (#3837) 2019-02-03 10:16:24 -08:00
Devin Petersohn
a1bcd2a4f5 Update Modin to 0.3.0 (#3936) 2019-02-02 23:06:16 -08:00
Michael Luo
1a015e420b Optimal PPO Configs (10k reward in 1 hr) + PPO grad clipping implemented (#3934) 2019-02-02 22:10:58 -08:00
Richard Liaw
eab6dd72b5
[tune] logging fixes, better warnings, better cluster support (#3906) 2019-02-02 19:14:03 -08:00
Luke
002531b199 Enable LZ4 compression in pyarrow build (#3931)
Enable LZ4 compression in pyarrow build
2019-02-02 14:38:02 -08:00
Yuhong Guo
54cbb4396f Prepare socket file when start ray (#3925) 2019-02-02 12:53:36 +08:00
Eric Liang
0f81bc9a33 [rllib] on_train_result results do not get logged (#3865) 2019-02-01 20:32:07 -08:00
Robert Nishihara
e0f82fd260 Fix building python 3.7 wheel by installing newer numpy. (#3927) 2019-02-01 18:06:48 -08:00
Daniel Edgecumbe
315edab085 [autoscaler] Speedups (#3720)
- NodeUpdater gets its' IP in parallel now (no longer in __init__)
- We use persistent connections in SSH (temp folder created only for ray; ControlMaster)
- hash_runtime_conf was performing a pointless hexlify step, wasting time on large files
- We use NodeUpdaterThreads and share the NodeProvider; NodeUpdaterProcess is removed
- AWSNodeProvider caches nodes more aggressively
- NodeProvider now has a shim batch terminate_nodes() call; AWSNodeProvider parallelises it; the autoscaler uses it
- AWSNodeProvider batches EC2 update_tags calls
- Logging changes throughout to provide standardised timing information for profiling
- Pulled out a few unnecessary is_running calls (NodeUpdater will loop waiting for SSH anyway)

## Related issue number
Issue #3599
2019-02-01 02:46:32 -08:00
Daniel Edgecumbe
ff3c6af1d6 [autoscaler]: Remove assertion in info string (#3916)
Fixes #3903
2019-02-01 00:32:24 -08:00
Tianming Xu
1302fafc0b [Tune] Add export_formats option to export policy graphs (#3868)
In earlier PRs, PR#3585 and PR#3637, export_policy_model and export_policy_checkpoint were introduced for users to export TensorFlow model and checkpoint.

For Ray Tune users, these APIs are not accessible through YAML configurations.

In this pull request, export_formats option is provided to enable users to choose the desired export format.
2019-01-31 17:07:27 -08:00
Kristian Hartikainen
b9eed2e86c [autoscaler] Move attach helper text under exec_cluster (#3920)
## What do these changes do?
Moves the attach command helper from cli commands to the actual `exec_cluster` function.
2019-01-31 17:01:24 -08:00
Peter Schafhalter
62a0a7bdc7 [tune] Add BayesOpt (#3864)
Adds BayesOpt as a Tune suggestion algorithm.
2019-01-31 16:54:17 -08:00