optimization

2002-04-21 Thread David Mechner
Hi, I have a question about how to optimize MySQL performance for a somewhat unusual (I think) usage pattern. In the first process, I'm summarizing a large amount of time series data (the main table will end up around 5GB, index around 1/3 of that) and storing the summaries to the db. During thi

Re[2]: delete where column <=> null

2002-11-19 Thread David Mechner
Ok, so it seems to be a bug. What's the procedure for reporting it (or do the developers read this list?) -David A minimal test case that demonstrates the bug: create table t ( id integer, index id_idx(id) ); insert into t(id) values(null); select * from t; delete from t where id <=> 1;

select from HEAP table failing with error 124

2001-01-15 Thread David Mechner
CREMENT, stock_idINTEGER NOT NULL REFERENCES stocks, data_type ENUM('LTG', 'EPS', 'REC', 'PRC') NOT NULL, periodicity ENUM('NONE', 'ANN', 'SEM', 'QTR') NOT NULL, period_yINTEGER, period_mINTEGER, unique index(stock_id,data_type,periodicity,period_y,period_m) ); What's going on? Thanks, -David Mechner