Re: MySQL 3.23.58 seg faults occasionally

2004-01-27 Thread Michael Bacarella
> I guess the table is corrupt and its row contains a wrong trx id. InnoDB > tries to retrieve an earlier version of the row, but the roll ptr in the row > is garbage too, and it contains a nonexistent rollback segment id. > > " > InnoDB: Error: trying to return an SQL NULL field in a non-null inno

Re: MySQL 3.23.58 seg faults occasionally

2004-01-27 Thread Heikki Tuuri
Michael, I guess the table is corrupt and its row contains a wrong trx id. InnoDB tries to retrieve an earlier version of the row, but the roll ptr in the row is garbage too, and it contains a nonexistent rollback segment id. " InnoDB: Error: trying to return an SQL NULL field in a non-null innoD

Re: MySQL 3.23.58 seg faults occasionally

2004-01-26 Thread Michael Bacarella
> > the crashes below happen in independent areas of code. The 2 first are > > inside InnoDB, and the third inside MySQL. This looks like random thread > > crashes, or random memory corruption. > > My colleague tells me that the third stack trace (the one inside MySQL) > is from a RESET MASTER. T

Re: MySQL 3.23.58 seg faults occasionally

2004-01-16 Thread Michael Bacarella
> the crashes below happen in independent areas of code. The 2 first are > inside InnoDB, and the third inside MySQL. This looks like random thread > crashes, or random memory corruption. My colleague tells me that the third stack trace (the one inside MySQL) is from a RESET MASTER. This server u

Re: MySQL 3.23.58 seg faults occasionally

2004-01-16 Thread Heikki Tuuri
Michael, the crashes below happen in independent areas of code. The 2 first are inside InnoDB, and the third inside MySQL. This looks like random thread crashes, or random memory corruption. I assume that you have my.cnf set so that the memory usage cannot approach 2 GB. You are running a relati