Re: dbload vs. load

> Is bullshit 'ClownSpeak' for RTFM? I have read the fu*** manual that is why i tried to insert the data with the load from statement. In the Manual there is a page with 4 questions about choosing the right import-utility. I had to answer those questions in that manner that i had choosing load as my result.... > IIRC, each operation (in this case, each INSERT) has to check or validate > something in the log, and does this in a relatively inefficient manner -- > traversing a linked list. So the longer the transaction gets, the slower > each subsequent operation is. When the transaction is short, the cost of > committing is relatively high, but when the transaction gets longer, it > can often be substantially better to commit more often (if this is > possible). I guess it is better to commit more often and therefore i have another question: i have tried many values for the commit-intervall and i never really found out what the max-value was. I tried with 10.000 and it didn't workout. I have informix sql version 9.40 > I hope I haven't cocked up the explanation too badly. > > -- > Bye now, > Obnoxio thank you for your help chris

christrier said: > > > > Is bullshit 'ClownSpeak' for RTFM? > I have read the fu*** manual that is why i tried to insert the data > with the load from statement. > In the Manual there is a page with 4 questions about choosing the right > import-utility. I had to answer those questions in that manner that i > had choosing load as my result.... > >> IIRC, each operation (in this case, each INSERT) has to check or >> validate >> something in the log, and does this in a relatively inefficient manner >> -- >> traversing a linked list. So the longer the transaction gets, the slower >> each subsequent operation is. When the transaction is short, the cost of >> committing is relatively high, but when the transaction gets longer, it >> can often be substantially better to commit more often (if this is >> possible). > > I guess it is better to commit more often and therefore i have another > question: i have tried many values for the commit-intervall and i never > really found out what the max-value was. I tried with 10.000 and it > didn't workout. I have informix sql version 9.40 How, exactly, did it not work out? One unspecified (by me, anyway) issue is the size of your logs and how much free log space you have. This ultimately constrains how big a transaction can be. But it's not really practical to say if "your logs are this big" and "your table is that wide" you can "stick so many rows" into a single transaction. If 10000 fails because of a long transaction, then try 1000. Or add a lot more logs. -- Bye now, Obnoxio "C'est pas parce qu'on n'a rien ' dire qu'il faut fermer sa gueule" - Coluche did i mention i like nulls? heck, i even go so far as to say that all columns in a table except the primary key could/should be nullable. this has certain advantages, for example, if you need to insert a child record and you don't have a parent row for it, just do an insert into the parent table with the primary key value (everything else null), and voila, relational integrity is preserved. but this is, admittedly, a bit controversial among modellers. --r937, dbforums.com