Bug #19382 | ndb_basic.test crashes/fails w. "Table definition has changed, please retry..." | ||
---|---|---|---|
Submitted: | 26 Apr 2006 19:01 | Modified: | 4 May 2006 21:50 |
Reporter: | Sergey Petrunya | Email Updates: | |
Status: | Duplicate | Impact on me: | |
Category: | MySQL Cluster: Cluster (NDB) storage engine | Severity: | S3 (Non-critical) |
Version: | 5.1-bk | OS: | |
Assigned to: | CPU Architecture: | Any |
[26 Apr 2006 19:01]
Sergey Petrunya
[26 Apr 2006 19:03]
Sergey Petrunya
The stacktrace from the core file: (I'll keep the tree+core file for some time, can look up more info if required) #0 0xb7d343e1 in kill () from /lib/libc.so.6 (gdb) wher #0 0xb7d343e1 in kill () from /lib/libc.so.6 #1 0xb7e92131 in pthread_kill () from /lib/libpthread.so.0 #2 0x08393ab9 in write_core (sig=6) at stacktrace.c:220 #3 0x0823a3c4 in handle_segfault (sig=6) at mysqld.cc:2148 #4 0xb7e94e55 in __pthread_sighandler () from /lib/libpthread.so.0 #5 <signal handler called> #6 0xb7d343e1 in kill () from /lib/libc.so.6 #7 0xb7e92131 in pthread_kill () from /lib/libpthread.so.0 #8 0xb7e924ab in raise () from /lib/libpthread.so.0 #9 0xb7d34174 in raise () from /lib/libc.so.6 #10 0xb7d3564d in abort () from /lib/libc.so.6 #11 0xb7d2d59f in __assert_fail () from /lib/libc.so.6 #12 0x085ba893 in Ndb::check_send_timeout (this=0x8b26048) at Ndbif.cpp:1060 #13 0x085ba6fb in Ndb::pollCompleted (this=0x8b26048, aCopyArray=0xb7d09034) at Ndbif.cpp:1036 #14 0x085bb0f9 in Ndb::poll_trans (this=0x8b26048, aMillisecondNumber=360000, minNoOfEventsToWakeup=1, pg=0xb7d0a054) at Ndbif.cpp:1330 #15 0x085bb06b in Ndb::sendPollNdb (this=0x8b26048, aMillisecondNumber=360000, minNoOfEventsToWakeup=1, forceSend=1) at Ndbif.cpp:1314 #16 0x085be7cf in NdbTransaction::executeNoBlobs (this=0x8b3bb18, aTypeOfExec=NoCommit, abortOption=AbortOnError, forceSend=1) at NdbTransaction.cpp:442 #17 0x085be269 in NdbTransaction::execute (this=0x8b3bb18, aTypeOfExec=NoCommit, abortOption=AbortOnError, forceSend=1) at NdbTransaction.cpp:273 #18 0x0840c228 in execute_no_commit (h=0x8b72300, trans=0x8b3bb18) at ha_ndbcluster.cc:269 #19 0x083f0042 in ha_ndbcluster::complemented_read (this=0x8b72300, old_data=0x8b731e0 "", new_data=0x8b731c0 "", old_part_id=0) at ha_ndbcluster.cc:1836 #20 0x083f2a48 in ha_ndbcluster::update_row (this=0x8b72300, old_data=0x8b731e0 "", new_data=0x8b731c0 "") at ha_ndbcluster.cc:2758 #21 0x0832fbf7 in handler::ha_update_row (this=0x8b72300, old_data=0x8b731e0 "", new_data=0x8b731c0 "") at handler.cc:3319 #22 0x082cc57f in mysql_update (thd=0xb51c7950, table_list=0x8b33fd8, fields=@0xb51c7c58, values=@0xb51c7e34, conds=0x0, order_num=0, order=0x0, limit=18446744073709551614, handle_duplicates=DUP_ERROR, ignore=false) at sql_update.cc:537 #23 0x08257689 in mysql_execute_command (thd=0xb51c7950) at sql_parse.cc:3230 #24 0x0825f0cd in mysql_parse (thd=0xb51c7950, inBuf=0x8b33f58 "UPDATE t1 SET pk1=pk1 + 1", length=25) at sql_parse.cc:5861 #25 0x082536f3 in dispatch_command (command=COM_QUERY, thd=0xb51c7950, packet=0xb51d8919 "UPDATE t1 SET pk1=pk1 + 1", packet_length=26) at sql_parse.cc:1736 #26 0x08252ed8 in do_command (thd=0xb51c7950) at sql_parse.cc:1532 #27 0x08251f4d in handle_one_connection (arg=0xb51c7950) at sql_parse.cc:1174 #28 0xb7e8f13d in pthread_start_thread () from /lib/libpthread.so.0 #29 0xb7dc11ba in clone () from /lib/libc.so.6
[27 Apr 2006 14:57]
Valeriy Kravchuk
The test passed OK for me on today's BK build: openxs@suse:~/dbs/5.1/mysql-test> ./mysql-test-run.pl --mysqld=--binlog-format=statement ndb_basic ... TEST RESULT TIME (ms) ------------------------------------------------------- ndb_basic [ pass ] 80178 ------------------------------------------------------- ... Will it fail for you when run alone?
[28 Apr 2006 7:03]
Sergey Petrunya
Valeriy, it will pass when run alone. Quoting the bug description: >>>> When I run ndb_basic.test alone it seems to always pass. Couple of tries with >>>> --star-from={several-tests-before-ndb_basic.test} also succeeded.
[28 Apr 2006 14:17]
Jim Winstead
I've seen the same test failure, too. I also get a failure of ndb_lock when I run the test suite with --force, but not when it is run alone.
[28 Apr 2006 16:02]
Valeriy Kravchuk
Yes, when run amond all other tests, ndb_basic failed for me also: ... ndb_basic [ fail ] Errors are (from /home/openxs/dbs/5.1/mysql-test/var/log/mysqltest-time) : mysqltest: At line 53: query 'UPDATE t1 SET attr1 = 9998 WHERE pk1 < 1000' failed: 1412: Table definition has changed, please retry transaction (the last lines may be the most important ones) Result from queries before failure can be found in r/ndb_basic.log Killing Possible Leftover Processes mysql-test-run: WARNING: can't kill old mysqld holding port 9310 Restoring snapshot of databases Resuming Tests On today's ChangeSet@1.2371, 2006-04-28 09:29:40+02:00, built using --with-debug=full.
[4 May 2006 21:50]
Tomas Ulin
bug 16445