Yi Jin created HAWQ-238:
---------------------------
Summary: Sending wrong size of message for returning YARN container
Key: HAWQ-238
URL: https://issues.apache.org/jira/browse/HAWQ-238
Project: Apache HAWQ
Issue Type: Bug
Components: Resource Manager
Reporter: Yi Jin
Assignee: Lei Chang
This causes resource manager hung when resource manager decides to return yarn
containers.
(lldb) bt
* thread #1: tid = 0x18f1d8, 0x938de736 libsystem_kernel.dylib`__read_nocancel
+ 10, queue = 'com.apple.main-thread', stop reason = signal SIGSTOP
* frame #0: 0x938de736 libsystem_kernel.dylib`__read_nocancel + 10
frame #1: 0x005fdf69 postgres`readPipe(fd=5, buff=0x03e07660, buffsize=16)
+ 68 at network_utils.c:602
frame #2: 0x00606706 postgres`handleRM2RB_ReturnResource + 289 at
resourcebroker_LIBYARN_proc.c:1011
frame #3: 0x0060461d postgres`ResBrokerMainInternal + 817 at
resourcebroker_LIBYARN_proc.c:257
frame #4: 0x006042aa postgres`ResBrokerMain + 434 at
resourcebroker_LIBYARN_proc.c:157
frame #5: 0x00601bd6 postgres`RB_LIBYARN_start(isforked='\x01') + 678 at
resourcebroker_LIBYARN.c:153
frame #6: 0x00600760 postgres`RB_start(isforked='\x01') + 51 at
resourcebroker_API.c:55
frame #7: 0x0063870d postgres`ResManagerMain(argc=3, argv=0xbffff418) +
1746 at resourcemanager.c:326
frame #8: 0x006389ea postgres`ResManagerProcessStartup + 428 at
resourcemanager.c:398
frame #9: 0x0040ead9 postgres`CommenceNormalOperations + 649 at
postmaster.c:3616
frame #10: 0x0040f6ef postgres`do_reaper + 2628 at postmaster.c:3964
frame #11: 0x0040ba0e postgres`ServerLoop + 984 at postmaster.c:2102
frame #12: 0x0040a81a postgres`PostmasterMain(argc=9, argv=0x038471e0) +
5127 at postmaster.c:1421
frame #13: 0x00321ed3 postgres`main(argc=9, argv=0x038471e0) + 1007 at
main.c:226
frame #14: 0x00002cbd postgres`_start + 212
frame #15: 0x00002be8 postgres`start + 40
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
