Hi, I tried to do updates on a Date field with an input flow which had 1000 rows on a table with 2138485 records. I used 2 ways: First way: tMap------------->tMMSQLOutput I used the 3 primary keys of my table to do the update. Result => this way is very slow, the job treat only 1 update per second Second way (dirty way): tMap------------->tFlowToIterate--------------->tMSSQLRow with the SQL request below
"UPDATE EEEA_DLVPLANSALESLINES SET ATW_DATE = GETDATE()" WHERE EEEA_DLVPLAN_ID ='"+(String)globalMap.get("EEEA_DLVPLAN_ID")+"' AND SALESID='"+(String)globalMap.get("SALESID")+"' AND DATAAREAID='"+(String)globalMap.get("DATAAREAID")+"'"
Result => this solution is 10x more faster. What I'm doing wrong with the tMSSQOutput component? Had someone already encountered this kind of trouble?
Hello together, talking about performance problems is not so easy if you give no detailed information about your job / infrastructure. And in the end we could only give you some hints or points to check but this will not supersede a detailed local analysis. Here some points to check / optimize: * Do you open the connection only one time? * Did you check your keys / indices? * Are there any network problems to access your db server? * If you need to insert big volumes, may it be possible to supress transactions or remove (temporary) indices? I hope this will help you a little bit. Bye Volker