On Mon, 2007-01-22 at 01:11 -0500, Jamie ffolliott wrote: > Hi Andrew, > > If it's going to mean more than say 10mins of potential downtime, I'm not in > a position to try experiments on the PDC server where this problem is > affecting us. > > I could try your patch on another linux box as a domain member server, would > that make sense, just to ensure it'll go smoothly on the PDC?
Are you seeing those symptoms on the member server? The fix I have isn't specific to a PDC however. > I'm attaching some valgrind output, I ran it briefly for about 15mins, it > was going slow but I managed to see some smb connections open to process > 26519 which was IUSR_JUNIPER, although I wasn't able to hit a webpage that > is accessed over the "web" share and handled by this pid. > > Attached is the valgrind output. > > The two PID's running that served iis5 websites are 26511 (boxwood IIS5 > server) and 26519 (juniper IIS5 server), and there's a little valgrind > output related to those pid's. > > Smbstatus showed, Thanks. I think the number of sessions explains some of the talloc dump. How big was the process we did the talloc pool dump of? (because 600k shouldn't kill a server). Andrew Bartlett -- Andrew Bartlett http://samba.org/~abartlet/ Authentication Developer, Samba Team http://samba.org Samba Developer, Red Hat Inc. http://redhat.com
signature.asc
Description: This is a digitally signed message part
