Re: [Proposal] Global temporary tables

Andrew Bille <andrewbille(at)gmail(dot)com> 于2021年10月28日周四 下午6:30写道：

> Thanks, the "group by" is fixed
>
> Yet another crash (on v58 patches), reproduced with:
>
> psql -t -c "create global temp table t(b text)
> with(on_commit_delete_rows=true); create index idx_b on t (b); insert into
> t values('test'); alter table t alter b type varchar;"
> server closed the connection unexpectedly
> This probably means the server terminated abnormally
> before or while processing the request.
> connection to server was lost
>
Thank you for pointing that out.
This is due to an optimization point: ALTER Table reuses the relfilenode of
the old index.
I have banned this optimization point for GTT, I am not entirely sure it is
appropriate, maybe you can give some suggestions.
Please review the new code(v59).

Wenjing

>
> with trace:
>
> [New LWP 569199]
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
> Core was generated by `postgres: andrew postgres [local] ALTER TABLE
> '.
> Program terminated with signal SIGABRT, Aborted.
> #0 __GI_raise (sig=sig(at)entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
> 50 ../sysdeps/unix/sysv/linux/raise.c: No such file or directory.
> (gdb) bt
> #0 __GI_raise (sig=sig(at)entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
> #1 0x00007f197493f859 in __GI_abort () at abort.c:79
> #2 0x00005562b3306fb9 in ExceptionalCondition
> (conditionName=0x5562b34dd740 "reln->md_num_open_segs[forkNum] == 0",
> errorType=0x5562b34dd72c "FailedAssertion", fileName=0x5562b34dd727 "md.c",
> lineNumber=187) at assert.c:69
> #3 0x00005562b3148f15 in mdcreate (reln=0x5562b41abdc0,
> forkNum=MAIN_FORKNUM, isRedo=false) at md.c:187
> #4 0x00005562b314b73f in smgrcreate (reln=0x5562b41abdc0,
> forknum=MAIN_FORKNUM, isRedo=false) at smgr.c:335
> #5 0x00005562b2d88b23 in RelationCreateStorage (rnode=...,
> relpersistence=103 'g', rel=0x7f196b597270) at storage.c:154
> #6 0x00005562b2d5a408 in index_build (heapRelation=0x7f196b58dc40,
> indexRelation=0x7f196b597270, indexInfo=0x5562b4167d60, isreindex=true,
> parallel=false) at index.c:3038
> #7 0x00005562b2d533c1 in RelationTruncateIndexes
> (heapRelation=0x7f196b58dc40, lockmode=1) at heap.c:3354
> #8 0x00005562b2d5360b in heap_truncate_one_rel (rel=0x7f196b58dc40) at
> heap.c:3452
> #9 0x00005562b2d53544 in heap_truncate (relids=0x5562b4167c58,
> is_global_temp=true) at heap.c:3410
> #10 0x00005562b2ea09fc in PreCommit_on_commit_actions () at
> tablecmds.c:16495
> #11 0x00005562b2d0d4ee in CommitTransaction () at xact.c:2140
> #12 0x00005562b2d0e320 in CommitTransactionCommand () at xact.c:2979
> #13 0x00005562b3151b7e in finish_xact_command () at postgres.c:2721
> #14 0x00005562b314f340 in exec_simple_query (query_string=0x5562b40c2170
> "create global temp table t(b text) with(on_commit_delete_rows=true);
> create index idx_b on t (b); insert into t values('test'); alter table t
> alter b type varchar;") at postgres.c:1239
> #15 0x00005562b3153f0a in PostgresMain (dbname=0x5562b40ed6e8 "postgres",
> username=0x5562b40ed6c8 "andrew") at postgres.c:4497
> #16 0x00005562b307df6e in BackendRun (port=0x5562b40e4500) at
> postmaster.c:4560
> #17 0x00005562b307d853 in BackendStartup (port=0x5562b40e4500) at
> postmaster.c:4288
> #18 0x00005562b3079a1d in ServerLoop () at postmaster.c:1801
> #19 0x00005562b30791b6 in PostmasterMain (argc=3, argv=0x5562b40bc5b0) at
> postmaster.c:1473
> #20 0x00005562b2f6d98e in main (argc=3, argv=0x5562b40bc5b0) at main.c:198
>
> On Mon, Oct 25, 2021 at 7:13 PM wenjing <wjzeng2012(at)gmail(dot)com> wrote:
>
>>
>> I missed whole row and system column. It has been fixed in v58.
>> Please review the new code(v58) again
>>
>>
>