Catalyst 6807 having parity error messages continue

[HCL Support]

Hi Team,

We are facing issue of the switch Cisco 6807 as the parity error coming continuously as the messages as below.

Jul 6 20:03:55.363 IST: SW2_DFC2: Warning: Uncorrectable Parity error is seen in the Netflow Table.
If any error is seen in Netflow flow creation or Netflow Data
export, it is advised to reboot the card in a maintenance window
Jul 6 20:03:55.383 IST: SW2_DFC2: Warning: Uncorrectable Parity error is seen in the Netflow Table.
If any error is seen in Netflow flow creation or Netflow Data
export, it is advised to reboot the card in a maintenance window
Jul 6 20:03:55.423 IST: %EARL-SW2_DFC2-1-EXCESSIVE_PARITY_ERROR: EARL 0: Parity error detected in VRAM

Please suggest if there is any Hardware part faulty or the software issue and what will be the permanent fix of the issue reported of the uncorrectable parity error detected in VRAM part.

Also attaching here the complete logs captured for this case the CLI show logging, show module, show diagnostic, show environment & suggest the solution on the case.

Regards,

Ashutosh

[Georg Pauwen]

Hello,

you might be running into this bug. Workaround is to reload the line card affected.

Excessive logging/Memory leak due to a parity error in the Netflow VRAM.
CSCus64795
Description
Symptom:
A Catalyst 6500 or 6800 may generate a lot of syslog messages and leak memory due to a parity error in the NetFlow table. A sample of these messages are:
%EARL-1-EXCESSIVE_PARITY_ERROR: EARL 0: Parity error detected in VRAM
and
Warning: Uncorrectable Parity error is seen in the Netflow Table.
If any error is seen in Netflow flow creation or Netflow Data
export, it is advised to reboot the card in a maintenance window

Conditions:
These messages are printed as a result of a parity error. This bug is not to address the parity error which has occurred in the system. Instead this bug was raised to overcome two issues:
1) The excessive debug logs that are produced by the system. From lab testing after a parity error the system will generate over 200,000 debug messages in a day.
2) The repeated misleading EXCESSIVE_PARITY_ERROR messages that get generated. From previous cases a single one time error will get flagged repeated by the system causing this misleading messages.
3) A memory leak that occurs as a result of the parity error. Each time the system flags the one bit error an entry into a table is created without bound. A single error will cause of a leak of roughly 30MB per day in the process named NF ISR Intr Task.

To stop the messages and leak schedule a window to reload the supervisor or linecard where the parity error was seen.

Workaround:
There is no workaround. The corrective action to clear the parity error is to perform a reload on the affected component in a maintenance window.

Further Problem Description:
Refer to CSCut56929 which addresses parity error in the VRAM.