Re: Rare crash

2001-07-30 Thread Michael Widenius
Hi! > "Sasha" == Sasha Pachev <[EMAIL PROTECTED]> writes: >> 2) select bigblob where optimized_index_clause order by >> optimized_index_clause limit 6650,50 >> >> Problem: I used to think that since the start number is high in the limit, >> that it was slow because the data was not often

Re: Rare crash

2001-07-27 Thread Sasha Pachev
Steve - thanks for positive feedback... > > We push MySQL to handle a lot of queries per second. Any query that last > several seconds can be enough to cause a cascade under a real heavy load. We > have three things that are slow: > > 1) select ... order by xyz desc > > Workaround: create an

Re: Rare crash

2001-07-20 Thread Sasha Pachev
On Friday 20 July 2001 14:00, Steven Roussey wrote: > > Locked table should not crash a server !! > > Oh.. I only meant in the cascade sense: Locked Table -> Queries waiting -> > Threads fill up -> Run out of memory -> Crash One thing to remember about row vs table level locking is that any que