Cisco 3750 - two ports stopped passing traffic

I have a WS-C3750V2-24PS-S   switch - two member stack - that has been up for four years.  It runs C3750-ADVIPSERVICESK9-M, Version 12.2(46)SE.  It's run flawlessly until the last few weeks.  Interface Gi1/0/6 stopped sending/receiving traffic although it was in up/up mode.  I tried to admin down/up the int but no good.  So I moved the cable to another interface (say 1/30) and the traffic flowed again.  Over the weekend interface Gi1/0/5 had the same symptom.  The host connected to it - a Cisco UCM stopped responding to ping or anything else although the port was up/up.  There were no errors visible when I tried "show log" and no incrementing errors on the interface, nothing in syslog.  I could even see the bps in and out appear to show some traffic.  But from a ping or TCP standpoint the interface was passing no traffic.  I bounced the UCM and still no traffic flowed.  I moved the cable to interface Gi 1/0/15 and now traffic was flowing again. 

I wanted to get some thoughts on here before opening a TAC case.  Does this sound like a memory leak bug with this version of IOS?  A failing ASIC?  Other thoughts on troubleshooting or possible explanation?

When traffic stops flowing, can you get any other device to communicate on those ports? Is any type of port security, storm control, dot1x, or any other method applied to the ports that could automatically disable a port?

Assuming that no security/error method (or spanning-tree) is blocking traffic, it does sound like an ASIC that is on the fritz. Perhaps a switch reboot is an order, and/or a TAC case. If you have smartnet, I would definitely start a TAC case immediately even if it's just a CYA. Grab a "show tech" when the issue is occurring for TAC  WS-C3750X-24P-S   to review.