Bug #17032 | ndb_autodiscover | ||
---|---|---|---|
Submitted: | 2 Feb 2006 8:19 | Modified: | 1 Jun 2006 11:37 |
Reporter: | Jonas Oreland | Email Updates: | |
Status: | Can't repeat | Impact on me: | |
Category: | MySQL Cluster: Cluster (NDB) storage engine | Severity: | S3 (Non-critical) |
Version: | 5.0 | OS: | Solaris (Solaris 10 AMD64 -max GCC) |
Assigned to: | Stewart Smith | CPU Architecture: | Any |
[2 Feb 2006 8:19]
Jonas Oreland
[3 Feb 2006 6:28]
Stewart Smith
I can see the failures in pushbuild. however, in my build on the machine, I am unable to reproduce with --do-test=ndb_autodiscover nor with running the whole test suite. I have run --do-test=ndb_autodiscover over 10 times and it has passed every one of them. am running full test suite again, but it has passed previously. maybe we need to engage with build team to get an exact replica of their build (i have something i think is close... but the exact environment is a bit of a mystery)
[3 Feb 2006 7:14]
Jonas Oreland
Look at end of printouts It creates a tar-ball of entire tree (built and everything) that you can use. That way you'll have the exact same build anyway
[16 Feb 2006 1:08]
Stewart Smith
Tried exact last failing build in a loop of running ndb_autodiscover. Out of 118 sequential tries, not one of them failed. Looking at the logs for the failing run, In /mysql-test/var/log/ndb_testrun.log an interesting thing during the ndb_autodiscover test we get the following: Dropping table 5... 709: No such table existed Dropping table UserTable... 709: No such table existed Dropping table Online... 709: No such table existed Dropping table Yes... 709: No such table existed Dropping table test... 709: No such table existed Dropping table def... 709: No such table existed Dropping table NDB$BLOB_4_1...OK NDBT_ProgramExit: 1 - Failed Could ndb_drop_table be being passed rubbish?
[20 Feb 2006 23:26]
Stewart Smith
seen again. Also on sparc. Now all I have to do is reliably reproduce it.