If you get a bunch of DUP ACKs, then you've got a layer 2 or layer 3 problem.
What happens with a continuous ping between the CAS and GC#2? Do packets get lost? From: [email protected] [mailto:[email protected]] On Behalf Of David Lum Sent: Tuesday, September 30, 2014 2:39 PM To: [email protected]; [email protected] Subject: [NTSysADM] LDAP requests never drop to zero Sorry for the cross-post, but this seems to fit both lists. Setup: I have a CAS box is looking at two GC's (single CAS, two mailbox servers). Most performance indicators on the CAs server fall well into the acceptable range, except for LDAP Search time and LDAP read time. LDAP read for GC #1: varies between 0 and 16ms. Avg = 3ms LDAP read for GC #2: varies between 0 and 3600ms. Avg = 1200ms LDAP search for GC #1: varies between 0 and 16ms. Avg = 4ms LDAP search for GC #2: varies between 267 and 289ms. Avg = 1240ms LDAP Outstanding requests for GC #1: varies between 0 and 1. LDAP Outstanding requests for GC #2: varies between 19 and 29, never dropping below 19 DCDIAG on GC#2 comes up clean. Nothing bizarre/unusual in the event logs. CPU, network and disk indicators on GC#2 show low utilization (or, at least not almost pegged). Doing a Network Monitor dump, I see a lot of "TCP: DUP ACK" between the CAS and GC #2 (about 10x more than between the CAS and the other GC), but I don't know if that's a the problem or the symptom. Anyone have a suggestion of other places for me to look, or should I look into the network? Notable is we recently changed CAS servers but we didn't have any baseline performance info before making the change, so I'm not sure if this issue existed beforehand or not. Dave Lum \\ I.T. Garage [email protected]<mailto:[email protected]> \\ 503.267.9764 (voice/text) www.theitgarage.com<http://www.theitgarage.com>
