Commit Graph

779 Commits

Author SHA1 Message Date
Tsuyoshi Ozawa 5a00b23106 YARN-4312. TestSubmitApplicationWithRMHA fails on branch-2.7 and branch-2.6 as some of the test cases time out. Contributed by Varun Saxena.
(cherry picked from commit 6636441911)
2015-11-05 10:40:23 -08:00
Sangjin Lee 6466ead9e0 Preparing for 2.6.3 development 2015-10-21 10:58:06 -07:00
Tsuyoshi Ozawa b898f8014f YARN-3798. ZKRMStateStore shouldn't create new session without occurrance of SESSIONEXPIED. (ozawa and Varun Saxena) 2015-10-21 23:08:02 +09:00
Jason Lowe ac865de725 YARN-3896. RMNode transitioned from RUNNING to REBOOTED because its response id has not been reset synchronously. (Jun Gong via rohithsharmaks)
(cherry picked from commit feaf034994)

Conflicts:

	hadoop-yarn-project/CHANGES.txt
2015-10-08 16:39:46 +00:00
Jason Lowe 528b809d2d YARN-3194. RM should handle NMContainerStatuses sent by NM while registering if NM is Reconnected node. Contributed by Rohith
(cherry picked from commit a64dd3d24b)
2015-10-08 16:33:34 +00:00
Jason Lowe 1484ebb602 YARN-3802. Two RMNodes for the same NodeId are used in RM sometimes
after NM is reconnected. Contributed by zhihai xu
(cherry picked from commit 5b5bb8dcdc)

Conflicts:

	hadoop-yarn-project/CHANGES.txt
2015-10-08 16:01:20 +00:00
Jason Lowe 4770f190b8 YARN-3780. Should use equals when compare Resource in
RMNodeImpl#ReconnectNodeTransition. Contributed by zhihai xu.
(cherry picked from commit c7ee6c151c)

Conflicts:

	hadoop-yarn-project/CHANGES.txt
2015-10-08 15:37:08 +00:00
Xuan 1828ba00be YARN-4087. Followup fixes after YARN-2019 regarding RM behavior when state-store error occurs. Contributed by Jian He
(cherry picked from commit 9f7fcb54e798cf4fda1ea7972dd96491976e1857)
2015-09-25 16:43:06 -07:00
Xuan d27f09c936 YARN-2019. Retrospect on decision of making RM crashed if any exception throw in ZKRMStateStore. Contributed by Jian He.
(cherry picked from commit db57d91ac91e895bcb9a23fa50af0b2fbcb1db5a)
2015-09-25 16:30:49 -07:00
Jian He c09bb46579 YARN-4101. RM should print alert messages if Zookeeper and Resourcemanager gets connection issue. Contributed by Xuan Gong
(cherry picked from commit 214fd1408c21f596d1d15217c11b58b34561aab7)
2015-09-25 16:26:18 -07:00
Jian He cc30002bc8 YARN-4092. Fixed UI redirection to print useful messages when both RMs are in standby mode. Contributed by Xuan Gong
(cherry picked from commit 6b3b487d3f4883a6e849c71886da52c4c4d9f0bf)
2015-09-25 16:25:13 -07:00
Sangjin Lee 4cb7dbaead Preparing for 2.6.2 development: mvn versions:set -DnewVersion=2.6.2 2015-09-25 15:51:13 -07:00
Zhijie Shen d57c3f0a26 YARN-3544. Got back AM logs link on the RM web UI for a completed app. Contributed by Xuan Gong.
(cherry picked from commit 21bf2cdcb77f69abc906e6cd401a8fb221f250e9)
(cherry picked from commit c9ee316045)
2015-09-15 17:30:15 -07:00
Xuan 7af5d6b4ba YARN-3248. Display count of nodes blacklisted by apps in the web UI.
Contributed by Varun Vasudev

(cherry picked from commit 4728bdfa15)
(cherry picked from commit e26b6e55e9)
2015-09-15 17:30:06 -07:00
Jian He e914220ab9 YARN-3379. Fixed missing data in localityTable and ResourceRequests table in RM WebUI. Contributed by Xuan Gong
(cherry picked from commit 4e886eb9cb)

(cherry picked from commit 3f0c9e5fe3)
2015-09-14 12:54:01 -07:00
Zhijie Shen f4154bdee8 YARN-1884. Added nodeHttpAddress into ContainerReport and fixed the link to NM web page. Contributed by Xuan Gong.
(cherry picked from commit 85f6d67fa7)
(cherry picked from commit 426535007b)
2015-09-11 11:45:29 -07:00
Vinod Kumar Vavilapalli 3462a00dd2 Preparing for release 2.6.1: mvn versions:set -DnewVersion=2.6.1 2015-09-09 15:29:57 -07:00
Jian He 2b526ba757 YARN-4047. ClientRMService getApplications has high scheduler lock contention. Contributed by Jason Lowe
(cherry picked from commit 7a445fcfab)

(cherry picked from commit 703fa1b141)
2015-09-08 22:57:35 -07:00
Xuan d59bf81e08 YARN-3999. RM hangs on draing events. Contributed by Jian He
(cherry picked from commit 3ae716fa69)
(cherry picked from commit 2ebdf5bfce)
2015-09-08 22:57:35 -07:00
Jonathan Eagles 6ed2486c7e YARN-3978. Configurably turn off the saving of container info in Generic AHS (Eric Payne via jeagles)
(cherry picked from commit 3cd02b9522)
(cherry picked from commit 899df5bce0)
2015-09-08 22:57:34 -07:00
Jian He 92742b4402 YARN-2301. Improved yarn container command. Contributed by Naganarasimha G R
(cherry picked from commit 258623ff8b)

(cherry picked from commit 1d1e7682c9)
2015-09-08 22:57:28 -07:00
Jian He 2336264900 YARN-2918. RM should not fail on startup if queue's configured labels do not exist in cluster-node-labels. Contributed by Wangda Tan
(cherry picked from commit f489a4ec96)

(cherry picked from commit d817fbb34d)
2015-09-06 14:15:33 -07:00
Jian He ee2b6bc248 YARN-3124. Fixed CS LeafQueue/ParentQueue to use QueueCapacities to track capacities-by-label. Contributed by Wangda Tan
(cherry picked from commit 18a594257e)

(cherry picked from commit 1be2d64ddd)
2015-09-06 11:54:40 -07:00
Jian He 637e7f9e39 YARN-2694. Ensure only single node label specified in ResourceRequest. Contributed by Wangda Tan
(cherry picked from commit c1957fef29)

(cherry picked from commit 3ddafaa7c8)
2015-09-05 21:07:51 -07:00
Jian He 4c94f07140 YARN-3098. Created common QueueCapacities class in Capacity Scheduler to track capacities-by-labels of queues. Contributed by Wangda Tan
(cherry picked from commit 21d80b3dd9)

(cherry picked from commit c0b1311a93)
2015-09-05 20:54:20 -07:00
Jian He d9281fbbab YARN-3099. Capacity Scheduler LeafQueue/ParentQueue should use ResourceUsage to track used-resources-by-label. Contributed by Wangda Tan
(cherry picked from commit 86358221fc)

(cherry picked from commit cabf97ae4f)
2015-09-05 20:54:20 -07:00
Jian He b0ad553841 YARN-3092. Created a common ResourceUsage class to track labeled resource usages in Capacity Scheduler. Contributed by Wangda Tan
(cherry picked from commit 6f9fe76918)

(cherry picked from commit 61b4116b4b)
2015-09-05 20:54:19 -07:00
Jian He 419e18cb37 YARN-2978. Fixed potential NPE while getting queue info. Contributed by Varun Saxena
(cherry picked from commit dd57c2047b)

(cherry picked from commit c61e8a7bfa)
2015-09-05 20:54:19 -07:00
Jian He 88f022da24 YARN-2920. Changed CapacityScheduler to kill containers on nodes where node labels are changed. Contributed by Wangda Tan
(cherry picked from commit fdf042dfff)

(cherry picked from commit 411836b74c)
2015-09-05 20:54:18 -07:00
Wangda Tan 85d92721a4 YARN-3733. Fix DominantRC#compare() does not work as expected if cluster resource is empty. (Rohith Sharmaks via wangda)
(cherry picked from commit ebd797c48f)
(cherry picked from commit 78d626fa89)
2015-09-03 17:43:01 -07:00
Jian He f1b35ffd4c YARN-2637. Fixed max-am-resource-percent calculation in CapacityScheduler when activating applications. Contributed by Craig Welch
(cherry picked from commit c53420f583)

(cherry picked from commit 4931600030)
2015-09-03 17:40:24 -07:00
Jason Lowe ca7fe71000 YARN-3990. AsyncDispatcher may overloaded with RMAppNodeUpdateEvent when Node is connected/disconnected. Contributed by Bibin A Chundatt
(cherry picked from commit 32e490b6c0)

(cherry picked from commit c31e3ba921)
(cherry picked from commit 07d31d4c0808a169f4770187d655f38aa105255c)
2015-09-03 14:40:20 -07:00
Xuan 7b1a71a7ad YARN-3526. ApplicationMaster tracking URL is incorrectly redirected on a QJM cluster. Contributed by Weiwei Yang
(cherry picked from commit b0ad644083)
(cherry picked from commit 802676e1be)
(cherry picked from commit 2cadeb9e017c6a75db16e1f23b2accda04f12298)
2015-09-03 11:54:23 -07:00
Wangda Tan e081593042 YARN-3487. CapacityScheduler scheduler lock obtained unnecessarily when calling getQueue (Jason Lowe via wangda)
(cherry picked from commit f47a5763ac)
(cherry picked from commit 3316cd4357)
(cherry picked from commit 24d45ee9544abcfcf9e611ab835ec2f824333670)
2015-09-02 11:28:22 -07:00
Wangda Tan 61f2ddb125 YARN-3493. RM fails to come up with error "Failed to load/recover state" when mem settings are changed. (Jian He via wangda)
(cherry picked from commit f65eeb412d)
(cherry picked from commit e7cbecddc3)
(cherry picked from commit 9d47d5aa5bffe427c4a77260f7ccc039d446e1fd)
2015-09-02 11:14:35 -07:00
Vinod Kumar Vavilapalli 752e3da738 YARN-3055. Fixed ResourceManager's DelegationTokenRenewer to not stop token renewal of applications part of a bigger workflow. Contributed by Daryn Sharp.
(cherry picked from commit 9c5911294e)
(cherry picked from commit 1ff3fd33ed)
(cherry picked from commit 82c722aae86669325672dd10840447434f15e7fd)
2015-09-01 21:31:00 -07:00
Wangda Tan 005d865494 YARN-3369. Missing NullPointer check in AppSchedulingInfo causes RM to die. (Brahma Reddy Battula via wangda)
(cherry picked from commit 6bc7710ec7)
(cherry picked from commit 8e142d27cb)
(cherry picked from commit 4d43be3c01b1bc0deb31a9081fca5395d0eb4e0d)
2015-09-01 17:10:42 -07:00
Jian He 7ffdf7d105 YARN-1809. Synchronize RM and TimeLineServer Web-UIs. Contributed by Zhijie Shen and Xuan Gong
(cherry picked from commit 95bfd087dc)

(cherry picked from commit a5f3fb4dc1)
(cherry picked from commit 27a2f0acb84202cc082090eef7eea57f6e42f9bb)
2015-09-01 15:12:53 -07:00
Tsuyoshi Ozawa 81417f7572 YARN-3249. Add a 'kill application' button to Resource Manager's Web UI. Contributed by Ryu Kobayashi.
(cherry picked from commit 1b67209612)
(cherry picked from commit 6660c2f83b)
(cherry picked from commit 6ea859e435e7cd6bc342f67e1551ccb86fbd976f)
2015-09-01 14:37:21 -07:00
Wangda Tan 8b5bdac98e YARN-3230. Clarify application states on the web UI. (Jian He via wangda)
(cherry picked from commit ce5bf927c3)
(cherry picked from commit a1963968d2)
(cherry picked from commit 591e261ccf1fb5dd25e87665c8d5c0341ff6fb24)
2015-09-01 14:34:07 -07:00
Karthik Kambatla 5a6755cc0f YARN-3242. Asynchrony in ZK-close can lead to ZKRMStateStore watcher receiving events for old client. (Zhihai Xu via kasha)
(cherry picked from commit 8d88691d16)
(cherry picked from commit 0d62e94887)
(cherry picked from commit 4a5b0e708d42fbff571229a43d1762d1767e2db5)
2015-09-01 14:06:34 -07:00
Karthik Kambatla dbc5bab9fd YARN-3231. FairScheduler: Changing queueMaxRunningApps interferes with pending jobs. (Siqi Li via kasha)
(cherry picked from commit 22426a1c9f)
(cherry picked from commit 721d7b5741)
(cherry picked from commit 5dfa25f22a989222e8b3d1013117b3350a48b2c5)
2015-09-01 13:54:04 -07:00
Jian He db92b09e03 YARN-3222. Fixed NPE on RMNodeImpl#ReconnectNodeTransition when a node is reconnected with a different port. Contributed by Rohith Sharmaks
(cherry picked from commit b2f1ec312e)

(cherry picked from commit 888a445638)
(cherry picked from commit b78f87825bd593e30b2f2ea76f37c7a4fd673ab2)
2015-09-01 13:39:35 -07:00
Zhijie Shen 28160a0bd6 YARN-2246. Made the proxy tracking URL always be http(s)://proxy addr:port/proxy/<appId> to avoid duplicate sections. Contributed by Devaraj K.
(cherry picked from commit d5855c0e46)
(cherry picked from commit fd75b8c9ca)
(cherry picked from commit a62891971380e5f8e4a645ed36bd88aa6fe0e47a)
2015-08-31 17:38:51 -07:00
Jian He a703952d39 YARN-3094. Reset timer for liveness monitors after RM recovery. Contributed by Jun Gong
(cherry picked from commit 0af6a99a3f)

(cherry picked from commit 6146680955)
(cherry picked from commit ab654746fbad2da12b24b13425dc9bf17c46b50c)
2015-08-31 17:17:47 -07:00
Tsuyoshi Ozawa 03f9ac2de7 YARN-2922. ConcurrentModificationException in CapacityScheduler's LeafQueue. Contributed by Rohith Sharmaks.
(cherry picked from commit ddc5be48fc)
(cherry picked from commit c116743bdd)
(cherry picked from commit 3c9d26ae14625de3e9437c07eceda0d05f1985b2)
2015-08-30 20:43:14 -07:00
Jian He e7fc071906 YARN-2992. ZKRMStateStore crashes due to session expiry. Contributed by Karthik Kambatla
(cherry picked from commit 1454efe5d4)

(cherry picked from commit ca0349b87a)
(cherry picked from commit 2f6be218fa41fd0f39633ec5ed0df6e0fa0f54b6)
2015-08-30 20:42:06 -07:00
Jian He 8cdcb792a1 YARN-2340. Fixed NPE when queue is stopped during RM restart. Contributed by Rohith Sharmaks
(cherry picked from commit 0d89859b51)

(cherry picked from commit 8b398a66ca)
(cherry picked from commit 37f1789f419fa741c2fbae9855549d405e856bff)
2015-08-30 20:40:09 -07:00
Jian He 6dfdcf094d YARN-2952. Fixed incorrect version check in StateStore. Contributed by Rohith Sharmaks
(cherry picked from commit 808cba3821)

(cherry picked from commit 9180d11b3b)
2015-08-30 20:34:14 -07:00
Jason Lowe 570d52e53c YARN-2964. RM prematurely cancels tokens for jobs that submit jobs (oozie). Contributed by Jian He
(cherry picked from commit 0402bada19)

(cherry picked from commit 173664d70f)
(cherry picked from commit 04e71db1ce9572ae0641234a02b7db5d174668fd)
2015-08-30 20:06:07 -07:00