Re: MySQL crashes

2010-06-21 Thread Ann W. Harrison
Charlene wrote: Anybody have any idea why MySQL would start to have this error message every 4 or so days at midnight: Do look at the logs, but if the error always occurs at midnight, you might look for a disk backup or compression program that's scheduled to run than and accesses the

Re: how things get messed up

2010-02-15 Thread Ann W. Harrison
Martijn Tonies wrote: For example, the Firebird DBMS stores (longer) Blob data not right there in the record, so whenever you don't request the blob (that is, not selecting it), it ignores it completely and it can go through the file quickly. As do most of the MySQL storage engines.

Re: how things get messed up

2010-02-12 Thread Ann W. Harrison
Martijn Tonies wrote: For example, the Firebird DBMS stores (longer) Blob data not right there in the record, so whenever you don't request the blob (that is, not selecting it), it ignores it completely and it can go through the file quickly. As do most of the MySQL storage engines. Cheers,

Re: MySQL 6.0.2-alpha and Falcon and possible bug/problem

2007-09-09 Thread Ann W. Harrison
Mathieu Bruneau wrote: I never tried the falcon engine, but could it be that mysql is running out of file descriptor ? That's less likely with Falcon than with engines that put each table and index in its own file. By default, Falcon tables share a single tablespace. Regards, Ann --

Re: MySQL 6.0.2-alpha and Falcon and possible bug/problem

2007-09-09 Thread Ann W. Harrison
Mariella Petrini wrote: ...MySQL 6.0.2 with Falcon on Linux with Debian 4. I have compiled the source code for 64 bit executable. The system is an Intel 2 cpus 4 cores each, with 8 GB of RAM. After having created approximately 8,500 empty tables mysqld server was still alive, you could

Re: Blob data

2007-07-03 Thread Ann W. Harrison
Rick James wrote: Instead I broke blobs into pieces, inserting them with a sequence number. Understanding the underlying problem, that still seems like an unnatural way to store pictures and documents. Added benefit: Does not clog up replication while huge single-insert is being copied