Supervisor Engine Is Faulty Due to NVRAM Failure or show version Command Gives Checksum Failures
If the Supervisor Engine shows the faulty status due to NVRAM component failure, as the show test command output in this section shows, and the show version command gives checksum failure as the command output shows, the problem can be Cisco bug ID CSCdx87646. The caveat is resolved in CatOS software release 7.5(1) or later.
VTP checksum failed.
!--- or
Global checksum failed.
Uptime is 8 days, 9 hours, 3 minutes
Console>
The global checksum error means that the next time the box is reloaded, the NVRAM most likely is lost due to a CRC checksum that failed while it read the configuration. This is not usually a hardware error, but the switch corrects itself. This does not have any impact to an operational switch unless configuration changes are made while the switch is in this condition. But most of the time, a reset resolves the checksum failure as it is re-calculated.
Complete these steps in order to recover the switch from this error state:
1. Backup the configuration of the switch. Refer to Uploading Configuration Files to a TFTP Server for more information on backing the configuration.
2. Reset the Supervisor module by issuing the reset supervisor_module_# command.
3. Once the switch boots up, issue the show version and show test commands in order to verify if the output is normal.
4. Verify the configuration existing on the switch, and restore from the backup if required.