[
https://issues.apache.org/jira/browse/HAWQ-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15944722#comment-15944722
]
Ming LI commented on HAWQ-1417:
-------------------------------
The root cause is: when ANALYZE is running, it will commit previous
transaction, and begin one transaction for each relation to process analyze.
However some code added before begining transactions, which means in this time,
no transaction, and no resource owner either.
The solution is to: create one resource owner for analyzeStmt(), and delete it
after processing it.
> Crashed at ANALYZE after COPY
> -----------------------------
>
> Key: HAWQ-1417
> URL: https://issues.apache.org/jira/browse/HAWQ-1417
> Project: Apache HAWQ
> Issue Type: Bug
> Components: Core
> Reporter: Ming LI
> Assignee: Ming LI
> Fix For: backlog
>
>
> This is the line in the master log where the PANIC is reported:
> {code}
> (gdb) bt
> #0 0x00007f6d35b0e6ab in raise () from
> /data/logs/52280/new_panic/packcore-core.postgres.457052/lib64/libpthread.so.0
> #1 0x00000000008c7d79 in SafeHandlerForSegvBusIll (postgres_signal_arg=11,
> processName=<optimized out>) at elog.c:4519
> #2 <signal handler called>
> #3 ResourceOwnerEnlargeRelationRefs (owner=0x0) at resowner.c:708
> #4 0x00000000008b5659 in RelationIncrementReferenceCount (rel=0x1baf500) at
> relcache.c:1941
> #5 RelationIdGetRelation (relationId=relationId@entry=1259) at
> relcache.c:1895
> #6 0x00000000004ca664 in relation_open (lockmode=lockmode@entry=1,
> relationId=relationId@entry=1259) at heapam.c:882
> #7 heap_open (relationId=relationId@entry=1259, lockmode=lockmode@entry=1)
> at heapam.c:1285
> #8 0x00000000008b0945 in ScanPgRelation (targetRelId=targetRelId@entry=5010,
> indexOK=indexOK@entry=1 '\001',
> pg_class_relation=pg_class_relation@entry=0x7ffdf2aed390) at relcache.c:279
> #9 0x00000000008b4302 in RelationBuildDesc (targetRelId=5010,
> insertIt=<optimized out>) at relcache.c:1209
> #10 0x00000000008b56c7 in RelationIdGetRelation
> (relationId=relationId@entry=5010) at relcache.c:1918
> #11 0x00000000004ca664 in relation_open (lockmode=<optimized out>,
> relationId=5010) at heapam.c:882
> #12 heap_open (relationId=5010, lockmode=<optimized out>) at heapam.c:1285
> #13 0x000000000055d1e6 in caql_basic_fn_all (pcql=0x1d70a58,
> bLockEntireTable=0 '\000', pCtx=0x7ffdf2aed480, pchn=0xf4b328
> <parser_cache+9608>) at caqlanalyze.c:343
> #14 caql_switch (pchn=pchn@entry=0xf4b328 <parser_cache+9608>,
> pCtx=pCtx@entry=0x7ffdf2aed480, pcql=pcql@entry=0x1d70a58) at
> caqlanalyze.c:229
> #15 0x00000000005636db in caql_getcount (pCtx0=pCtx0@entry=0x0,
> pcql=0x1d70a58) at caqlaccess.c:367
> #16 0x00000000009ddc47 in rel_is_partitioned (relid=1882211) at
> cdbpartition.c:232
> #17 rel_part_status (relid=relid@entry=1882211) at cdbpartition.c:484
> #18 0x00000000005e7d43 in calculate_virtual_segment_number
> (candidateOids=<optimized out>) at analyze.c:833
> #19 analyzeStmt (stmt=stmt@entry=0x2045dd0, relids=relids@entry=0x0,
> preferred_seg_num=preferred_seg_num@entry=-1) at analyze.c:486
> #20 0x00000000005e89a7 in analyzeStatement (stmt=stmt@entry=0x2045dd0,
> relids=relids@entry=0x0, preferred_seg_num=preferred_seg_num@entry=-1) at
> analyze.c:271
> #21 0x000000000065c25c in vacuum (vacstmt=vacstmt@entry=0x2045bf0,
> relids=relids@entry=0x0, preferred_seg_num=preferred_seg_num@entry=-1) at
> vacuum.c:316
> #22 0x00000000007f6012 in ProcessUtility
> (parsetree=parsetree@entry=0x2045bf0, queryString=0x2045d30 "ANALYZE
> mis_data_ig_account_details.e_event_1_0_102", params=0x0,
> isTopLevel=isTopLevel@entry=1 '\001', dest=dest@entry=0xf04ba0 <donothingDR>,
> completionTag=completionTag@entry=0x7ffdf2aee3f0 "") at utility.c:1471
> #23 0x00000000007f1ade in PortalRunUtility (portal=portal@entry=0x1bfb490,
> utilityStmt=utilityStmt@entry=0x2045bf0, isTopLevel=isTopLevel@entry=1
> '\001', dest=dest@entry=0xf04ba0 <donothingDR>,
> completionTag=completionTag@entry=0x7ffdf2aee3f0 "") at pquery.c:1968
> #24 0x00000000007f32be in PortalRunMulti (portal=portal@entry=0x1bfb490,
> isTopLevel=isTopLevel@entry=1 '\001', dest=0xf04ba0 <donothingDR>,
> dest@entry=0x1b93f60, altdest=0xf04ba0 <donothingDR>,
> altdest@entry=0x1b93f60, completionTag=completionTag@entry=0x7ffdf2aee3f0 "")
> at pquery.c:2078
> #25 0x00000000007f5025 in PortalRun (portal=portal@entry=0x1bfb490,
> count=count@entry=9223372036854775807, isTopLevel=isTopLevel@entry=1 '\001',
> dest=dest@entry=0x1b93f60, altdest=altdest@entry=0x1b93f60,
> completionTag=completionTag@entry=0x7ffdf2aee3f0 "") at pquery.c:1595
> #26 0x00000000007ee098 in exec_execute_message (max_rows=9223372036854775807,
> portal_name=0x1b93ad0 "") at postgres.c:2782
> #27 PostgresMain (argc=<optimized out>, argv=<optimized out>,
> argv@entry=0x1a49b40, username=0x1a498f0 "mis_ig") at postgres.c:5170
> #28 0x00000000007a0390 in BackendRun (port=0x1a185f0) at postmaster.c:5915
> #29 BackendStartup (port=0x1a185f0) at postmaster.c:5484
> #30 ServerLoop () at postmaster.c:2163
> #31 0x00000000007a3159 in PostmasterMain (argc=<optimized out>,
> argv=<optimized out>) at postmaster.c:1454
> #32 0x00000000004a52b9 in main (argc=9, argv=0x1a20d10) at main.c:226
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
