[BUG] possibly in tdb

[Luke Kenneth Casson Leighton]

On Mon, 14 Feb 2000, Luke Kenneth Casson Leighton wrote:

> andrew, yesterday's rpctorture test had 50 simultaneous processes making
> 100 sequential logins to a single server.  number of processes went up
> from 30 to 185, and pretty much stayed there.  loadavg up to 5-7.

> i'll try the torture test again and see if i can repro this.

ok, this time i get between 160 and 200 processes (fluctuating).  this of
course includes the 50 rpctorture processes themselves.  loadavg is up to
30-40, and i immediately have one of the login test processes failing
consistently.

after this test run, i'll met you know if a simple login works.

ok, about 1/3 way through, a whole batch has now failed.  i'm down to 127
processes because there are only 25 rpctorture processes still running.
loadavg hit 40 at the time, and is now back down to 30.

that one process keeps saying no, no, no :) it's quite amusing :) at least
it's consistent!

dayum, things are running sloow today...

ok, an rpcclient ntlogin test in the _middle_ of this didn't fail, it must
be some sort of compounded problem.  hmm...

i wonder if i managed to kill netlogond in the middle of a tdb access,
would that be a problem, do you think?