We have a remote site, fiber-connected with a single 24-port powered switch. I am getting alot of reports of phones spontaeously resetting. When I look at the CCM Syslog, I see the reason code as Keepalive timeouts. However, QoS is configured on this switch and the phone resets are happening one at a time, rather than several at once. Even more bizarre, when I started looking for patterns, I found that only phones that had pulled an odd DHCP IP address are having a problem. No phone with an even numbered IP address has had a problem and every phone with an odd numbered IP has, typically 3 or 4 times a day.
The switch is a 3560-24PS running IPBase 12.2(46). There are no ACL's or firewalls on the voice network. Has anyone seen anything like this before?
I am currently waiting on one of the network engineers to get on-site to manually switch one of the phones to an even address for one of the power users to verify.