Active SUP in Cat6500 giving Centauri Interrupt L3 parity errors

[usabnis]

Ever since the upgrade from CatOS 5.5.7 to 5.5.13a

I am seeing continous errors on the Active SUP.

I did switch over to the standby SUPERVISOR during the upgrade. Anyone out there seen these errors ?

The errors are:

CentauriInterrupt:L3 parity Error 7f51cb 8001 8037743

titan_intr_hdlr:Last titan interrupt occurred at 00:00:00.0

titan_intr_hdlr: Main Intr 0x0; Module 0x0, value 0x0; fatal intrs since last read 0, last Src Index seen: 0x0

[bsivasub]

If you are seeing it continously, it indicates bad PFC. You need to have the sup replaced.

[usabnis]

Thank you for replying. I did go ahead and replace the SUP.

However can you tell me how these messages can translate into performance problems ?

Throughout the week that I had the errors I had been monitoring the L3 switching ( MLS ) tables and they seemed to be running fine.

SO what does a bad PFC mean ?

[bsivasub]

PFC contains the MLS short cuts programmed for hardware forwarding. So these indicated that software has detected parity error while reading/writing the flows. Depending on how much memory section has gone bad, you may/may not see any effect. Now if you are seeing this message on the redundant supervisor as If HA is enabled, the second PFC is also programmed but not used for forwarding of traffic. So it may not any impact for the forwarding.