Is it possible that there is some limitation to the number of statements
in a single transaction? I'm trying to debug a problem where a program
locks up during an insert[*].
Actually, there are 3 simultaneous transactions going on, all doing
inserts of about 3000 records each. The problem occurs when I'm about
400 records from the end of the transaction.
As they're inserts, and therefore not even touching the same data, I'm
quite certain it's not some kind of row locking issue (does that even
happen at all with MVCC?).
I'm just fishing around, it may just as well turn out to be a threading
problem in my program, but I've been staring at it for days now and it
doesn't seem to make more sense then it did at the begin...
Any insights much appreciated.
[*]: It prints debug output before the query is executed, and it should
after.
BTW: My thanks to whoever changed the color of the links in the docs to
the current color. The contrast/readability is much better now (compared
to last week).