Cisco 3640 w/ ATM Corrupt Packet Headers

[ccarr]

I have a 3640 w/ an ATM module, pvc config with sub-ints like a2/0.1, a2/0.2, etc...&nbsp;&nbsp;I have these sub-ints bound to virtual-access ints.&nbsp;&nbsp;Problem is I am constantly getting errors like <b><font color=red>(%IPFAST-2-PAKSTICK: Corrupted pak header for Virtual-Access2, flags 0x80)</font></b>.&nbsp;&nbsp;I've been told by cisco that this is an &quot;IP Fast Switching error while running ppp over ATM&quot;.&nbsp;&nbsp;Anyone know anything about this?&nbsp;&nbsp;or have any solutions?

 

[Guest_imported]

Try&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;no ip directed-broadcast<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;no ip route-cache cef<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;no ip route-cache distributed<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;no ip route-cache<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;no ip mroute-cache<br><br>under each sub-interface including any bvi's.<br><br>regards,<br>Rick<br>

 

[Guest_imported]

Just opened a case with Cisco with similar messages. This is on a 1720 running 12.0.5 with a full T1 on S0. They debugged the message and informed me that the router was acually crashing and need to go to 12.0.7. He could not tell me if this was across all hardware platforms or just the 1700s. <br>Another symptom was the was would go up and down - power cycling the box seemed to take care of the problem for a while.<br>Don't know if this helps or applies - but that was my story.<br>Good Luck