On Tue, 26 Jul 2005 11:37:29 +0200, Pavel Cahyna wrote:

> I'm afraid that even if a steady stream was achieved, there would be
> exactly the same waste of CPU time, because ioflush would keep the vnode
> locked for the same amount of time.
> 
> Is my analysis correct?

My apologies: I realized that in the case discussed here where everything
(not just the process using this database) stops, my analysis don't apply
and the problem is probably really that RAM gets filled.

The lock contention is probably also a problem, but different. Maybe it
could explain why "data processing program is anyhow blocked until ioflush
finishes and can't do any useful work".

Bye	Pavel