what wrong about my 7507

[admin_2]

a pos interface of my 7507, it's line protocol has become "up" and "down" for some minutes.

This is the show log text:

Apr 14 14:41:17 cst: %RSP-3-ERROR: MD error 0000008000002000

-Traceback= 60332754 60332E70 60332FA8 603339A4 6030D3AC

Apr 14 14:41:17 cst: %RSP-3-ERROR: SRAM parity error (bytes 0:7) 02

-Traceback= 60332830 60332E70 60332FA8 603339A4 6030D3AC

Apr 14 14:41:17 cst: %VIP4-80 RM7000-3-MSG: slot5 VIP-3-MVIP_CYBUSERROR_INTERRUPT: A Cybus Error occurred.

Apr 14 14:41:18 cst: %VIP4-80 RM7000-3-MSG: slot5 VIP-3-SVIP_RELOAD: SVIP Reload is called.

Apr 14 14:41:18 cst: %VIP4-80 RM7000-3-MSG: slot5 VIP-3-SYSTEM_EXCEPTION: VIP System Exception occurred sig=22,

code=0x0, context=0x605FE788

Apr 14 14:41:19 cst: %DBUS-3-CXBUSERR: Slot 5, CBus Error

Apr 14 14:41:19 cst: %DBUS-3-DBUSINTERRSWSET: Slot 5, Internal Error due to VIP crash

Apr 14 14:41:19 cst: %RSP-3-ERROR: End of MEMD error interrupt processing

-Traceback= 60332F40 60332FA8 603339A4 6030D3AC

Apr 14 14:41:19 cst: %DBUS-3-CXBUSERR: Slot 5, CBus Error

Apr 14 14:41:19 cst: %DBUS-3-DBUSINTERRSWSET: Slot 5, Internal Error due to VIP crash

Apr 14 14:41:20 cst: %CBUS-3-CMDTIMEOUT: Cmd timed out, CCB 0x5800FF30, slot 1, cmd code 2

-Traceback= 6037B100 6037B5EC 603734A4 603714DC 602BE770 602BE9A0 602E0FC4 602E0FB0

Apr 14 14:41:20 cst: %CBUS-3-CMDTIMEOUT: Cmd timed out, CCB 0x5800FF60, slot 4, cmd code 2

-Traceback= 6037B100 6037B5EC 603734A4 603714DC 602BE770 602BE9A0 602E0FC4 602E0FB0

Apr 14 14:41:21 cst: %CBUS-3-CMDTIMEOUT: Cmd timed out, CCB 0x5800FF20, slot 0, cmd code 2

-Traceback= 6037B100 6037B5EC 603734A4 603714DC 602BE770 602BE9A0 602E0FC4 602E0FB0

Apr 14 14:41:21 cst: %IPC_RSP_CBUS-3-NOBUF: No more IPC memd buffers to transmit IPC message

Apr 14 14:41:22 cst: %LINK-3-UPDOWN: Interface GigabitEthernet1/0/0, changed state to down

[gfullage]

Please review the following page for troubleshooting IPC error messages :

http://www.cisco.com/warp/public/63/ipcnobuff_21165.html

Distributed CEF uses IPC (inter process communications) to update CEF table on the VIPS.

In some occasions, where a lot of routing updates are received or when we have to delete

an important amount of routes, we may be running

short on IPC buffers causing distributed CEF to be disabled on the linecards. It has to reenabled

again via "ip distributed cef". The number of IPC buffers has been made tunable to deal with that

situation.

I would recommend you configure : "ipc cache 5000" in global configuration mode and you should

get rid of this problem.

Having said that, you're also getting a traceback which implies a software bug, but I can't decode it without a "sho ver" output. I would suggest that if increasing the ipc buffers doesn't help, open a TAC case and include a "sho tech" output from this router. They'll be able to figure it out then.