Commit graph

9069 commits

Author SHA1 Message Date
Thomas Desrosiers
3e48df89f7
[Client] Fix mismatched debug log ID formats (#17597) 2021-08-13 13:28:20 -07:00
Huaiwei Sun
14365e111d
Update README.rst (#17471)
Add Slack channel info in the "Getting Involved" section
2021-08-13 13:25:51 -07:00
akern40
0cb2c602db
[rllib] Fixes typo in RolloutWorker.__init__ (#17583)
Fixes the typo in RolloutWorker.__init__, closes #17582
2021-08-13 13:17:36 -07:00
Amog Kamsetty
9f5dc5ec9f
[Docker] Downgrade to CUDA 11.0 (#17806) 2021-08-13 20:39:06 +02:00
architkulkarni
fcac416933
[Serve] [Dashboard] Add start times and replica tags to cluster snapshot (#17749) 2021-08-13 09:49:12 -07:00
Simon Mo
7d482fe099
[Doc] Update macos nightly wheel names (#17813) 2021-08-13 09:45:10 -07:00
Eric Liang
7ec52ca311
Make the namespace argument explicit instead of implicit in actor names (#17758) 2021-08-13 09:24:13 -07:00
Hasan Genc
a33dc75105
Shutdown clusters when large number of nodes (#17642)
* Allow clusters with over 1000 nodes to be shut down

* Add unit-test for terminating large number of nodes on AWS

* Fix lint

* Add max_terminate_nodes to the NodeProvider abstract class, and refactor terminate_nodes to reduce repetition

* lint

* Update comment

Co-authored-by: Ameer Haj Ali <ameerh@berkeley.edu>

* lint

* lint

* Unit test previously required internet access. This commit removes that requirement.

Co-authored-by: Ameer Haj Ali <ameerh@berkeley.edu>
2021-08-13 17:09:19 +03:00
Kai Fricke
96b620bc01
[docker] Pin matplotlib, fix docker build (#17819) 2021-08-13 14:59:50 +01:00
Qing Wang
9d5c68ff55
[Java] Better log message when failed to invoke task. (#17737) 2021-08-13 17:31:58 +08:00
xwjiang2010
0be9f06ab6
[tune] Output insufficent resources warning msg when trials are in pending for extended amount of time. (#17533) 2021-08-13 01:37:56 -07:00
Hao Zhang
61de23cbae
[Collective] silent the pygloo warning as it is not commonly used (#17792) 2021-08-13 00:47:45 -07:00
qicosmos
a2a1c46c83
[C++ Worker]Fix for mac (#17633)
* linkopts shared

* replace gflags with absl flags

* fix

* add test option

* fix

* add cpp worker to mac ci

* fix

* support empty redis password;mod arc argv

* add encoding

* test

* ignore example test on mac

* support mac

* fix

* fix and update doc

* fix

* fix run.sh

* fix init

* fix typo

* fix run.sh

* fix lint

Co-authored-by: 久龙 <guyang.sgy@antfin.com>
2021-08-13 12:22:37 +08:00
SangBin Cho
21635b32e5
[Core] Fix the segfault (#17772) 2021-08-12 18:17:50 -07:00
Simon Mo
242a5d1a8d
[Serve] Add support for root_url (#17765) 2021-08-12 17:54:53 -07:00
Simon Mo
22b030d79f
[Serve] Remove serve.start(http_*) arguments (#17762) 2021-08-12 17:50:12 -07:00
Guyang Song
b97027ec64
[C++ API] support cpu gpu num 0 (#17783)
* support cpu gpu num 0

* support cpu gpu num 0

* fix
2021-08-13 08:45:33 +08:00
Robert Nishihara
f624ddae5f
Remove outdated link from readme (#17788) 2021-08-12 17:11:59 -07:00
Yi Cheng
e32d33f39c
Fix ray.init hanging due to failure. (#17732)
* up

* change to 30s

* up

* up

* format
2021-08-12 16:56:10 -07:00
wanxing
e4c8125c86
Make some function private (#17729)
* ReceiveObjectChunk

* more
2021-08-12 15:27:37 -07:00
Eric Liang
7fc62a1529
Support dataset union (#17793) 2021-08-12 14:01:40 -07:00
Lixin Wei
d287fc941b
[Core] Add Running Count to instrumented_io_context (#17664) 2021-08-12 13:56:40 -07:00
Chen Shen
9565fa549e
[Core][RFC] limit the total number of inlined bytes in task request rpc
Co-authored-by: Clark Zinzow <clarkzinzow@gmail.com>
2021-08-12 13:55:54 -07:00
Simon Mo
ec8409ff06
Add @architkulkarni to snapshot code owner (#17785) 2021-08-12 10:58:02 -07:00
Simon Mo
6879293b6b
[CI] Mark some tests exclusive (#17650) 2021-08-12 10:28:03 -07:00
Guyang Song
88b8de5904
[C++ API] support ray::IsInitialized (#17780)
* support ray::IsInitialized

* address comments

* fix
2021-08-13 00:51:26 +08:00
SangBin Cho
8fd7e025be
Skip raylet kill windows #17682 (#17683)
* Try fixing it?

* Done

* skip raylet signal
2021-08-12 09:35:44 -07:00
matthewdeng
55680a1f9e
[SGD] v2 initial checkpoint functionality (#17632)
* [SGD] initial checkpoint functionality

* remove thread implementation and merge with fetch_next_result

* Update comment

Co-authored-by: Amog Kamsetty <amogkam@users.noreply.github.com>

* address comments

* add additional tests

* fix imports

* load most recently saved checkpoint

Co-authored-by: Amog Kamsetty <amogkam@users.noreply.github.com>
2021-08-12 08:52:04 -07:00
Clark Zinzow
d6eeb5dc70
[Datasets] Add local and S3 filesystem test coverage for file-based datasources. (#17158) 2021-08-12 08:39:31 -07:00
Guyang Song
e53aeca6bb
[C++ API]support set resources in RayConfig (#17779) 2021-08-12 22:53:42 +08:00
Guyang Song
5713a0be6c
[C++ API] add C++ API docs (#17743) 2021-08-12 22:40:09 +08:00
mguarin0
3e010c5760
[rllib] bug fix for rllib pettingzoo pistonball_v4 example (#17701)
* bug fix for rllib pettingzoo pistonball_v4 example

* adding test for PR 17701

* ran scripts/format.sh

* ok

Signed-off-by: Richard Liaw <rliaw@berkeley.edu>

Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2021-08-12 00:25:00 -07:00
architkulkarni
00f6b30684
[Serve] [Dashboard] Support nondetached and multiple Serve instances in cluster snapshot (#17747) 2021-08-11 22:26:54 -05:00
Eric Liang
ce171f10a1
Remove legacy plasma unlimited and pull manager pinning flag (#17753) 2021-08-11 20:19:12 -07:00
Guyang Song
63f9ba2858
[C++ API][Fix] support ray::Init without RayConfig (#17733) 2021-08-12 10:59:21 +08:00
Kai Yang
ab53c5fc93
[Java] Update rolling logging configuration (#17741) 2021-08-12 10:15:27 +08:00
Qing Wang
6d6a1ea43e
Support reading system configs from native in Java. (#17703)
* Support reading system configs from native in Java.

* Fix lint

* Lint cpp

* Fix Java cases.

* Address comments.

* Address comments.
2021-08-12 10:06:01 +08:00
Clark Zinzow
623db7c47b
[Datasets] Add support for reading partitioned Parquet datasets. (#17716) 2021-08-11 15:55:49 -07:00
Jiao
3c64a1a3c1
Add micro benchmark to releaser repo (#17727) 2021-08-11 15:15:33 -07:00
architkulkarni
9a70e83e90
[hotfix] pin tensorflow==2.5.1 (#17760)
* pin tensorflow==1.5.1

* Update python/requirements.txt

Co-authored-by: Richard Liaw <rliaw@berkeley.edu>
2021-08-11 15:15:22 -07:00
Yi Cheng
aa96e59faf
[workflow] Examples of function chaining (#17715) 2021-08-11 13:15:51 -07:00
Eric Liang
71b3183038
Add implicit init note to Ray docs & dataset version note (#17751) 2021-08-11 13:13:22 -07:00
Yi Cheng
02e79f3fe5
Revert "[Observability] Export useful metrics (#17578)" (#17752)
This reverts commit bd4db53df2.
2021-08-11 12:21:50 -07:00
Jiao
e38db5875b
Add serve external kv store (#17622) 2021-08-11 12:06:14 -07:00
Amog Kamsetty
ed24bae644
[SGD] Fail if num_workers is not greater than 0 (#17723) 2021-08-11 10:05:19 -07:00
Ian Rodney
97f7ae5e06
[Cluster Launcher] Allow attach/exec on uninitialized head node (#17688) 2021-08-11 09:43:23 -07:00
Sven Mika
7f2b3c0824
[RLlib] Issue 17667: CQL-torch + GPU not working (due to simple_optimizer=False; must use simple optimizer!). (#17742) 2021-08-11 18:30:21 +02:00
chenk008
f0fc26960d
[sgd] Wait for placement_group deletion when shutdown worker_group (#17698)
* fix

* fix ut

* delete sleep

* fix according to comment

* fix according to comment

* use pg in test_resize

* fix
2021-08-11 08:47:49 -07:00
Tricia Fu
24c4220bd7
[doc][serve] Update http-servehandle.rst (#17680) 2021-08-11 10:39:58 -05:00
Julius Frost
6891dee6ea
[RLlib] Better exceptions with traceback in TorchPolicy (#17690) 2021-08-11 15:01:07 +02:00