Jerry Amundson wrote:

Server's Redhat 9 (outdated, I know...that upgrade's on the to-do list, too heh). As far as I know it was kept reasonably updated up until about a month ago when the previous admin left.

The clients are primarily thunderbird, but some users use Outlook Express. All clients encounter the same delay.

Checking with top the only thing I noticed that was high was a high amount of cached memory (1.1Gigs as I check it this morning), but I've seen similar usage, especially on mail servers that do a lot of content scanning, so I didn't really think to mention it. I'll paste what top says right now, but the delay's not happening right now so it's not likely to be of much use other than as a baseline.

09:03:31 up 24 days, 16:06, 5 users, load average: 0.34, 0.17, 0.10 169 processes: 165 sleeping, 3 running, 1 zombie, 0 stopped CPU states: 0.2% user 1.4% system 0.0% nice 0.0% iowait 98.3% idle Mem: 2048132k av, 1967496k used, 80636k free, 0k shrd, 251036k buff 1395236k actv, 118536k in_d, 40512k in_c Swap: 2040244k av, 215792k used, 1824452k free 1149252k cached

I've also sniffed with ethereal to see if it was a high volume of traffic (a daily hit from a big spammer, or something), and to verify that it's not doing dns lookups (also used tcpdump to check for this and didn't see anything really notable, either), and there wasn't anything odd about the amount or nature of the traffic.

Sam Varshavchik mentioned the maximum amount of simultaneous connections, and that'd been something I'd decided to increase as one of my tests. I'll post again if it doesn't help the problem.

Bob..