From: Mark G. <mg...@us...> - 2005-06-30 19:49:49
|
Update of /cvsroot/gmod/apollo/doc/chado In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv9779/doc/chado Modified Files: transaction-todo Log Message: with the new much bigger rice database, inserting data ground to a halt - it was taking over 10 minutes to insert a simple little gene. theres 2 issues here - the main one is select max id from feature - to find out what id got inserted. the max id query takes over a minute and is done for every feature (gene,trans,prot,ex). its really odd that it takes so long as there is an index on it so you would think it would just shoot to the end of the index. so i put in the nextval function which figures the next id to be inserted. and then does the insert with this id. the bummer is that i imagine nextval is postgres specific - this will need to be configured when other vendors want to use writeback. the other slowness is feat_rel insert which must be a trigger issue - scott is looking into it. Index: transaction-todo =================================================================== RCS file: /cvsroot/gmod/apollo/doc/chado/transaction-todo,v retrieving revision 1.137 retrieving revision 1.138 diff -C2 -d -r1.137 -r1.138 *** transaction-todo 30 Jun 2005 13:46:59 -0000 1.137 --- transaction-todo 30 Jun 2005 19:49:40 -0000 1.138 *************** *** 212,215 **** --- 212,218 ---- check for redundant syn names and drop them? + would be nice if could select what features to get in adapter gui - + predictions, hits, genes... + with new db dump switch protein to polypeptide - make configurable need to also change chado transaction output and template xml file |