Developers Guide

Fabric OS Troubleshooting and Diagnostics Guide 15
53-1003141-01
Switch boot
2
• Software Fault: Kernel Panic
- If the system detects an internal fatal error from which it cannot safely recover, it outputs
an error message to the console, dumps a stack trace for debugging, and then performs
an automatic reboot.
- After a kernel panic, the system may not have enough time to write the reboot reason,
causing the reboot reason to be empty. This is treated as a reset case.
• Software fault
- Software Watchdog
- ASSERT
• Software recovery failure
This is an HA bootup-related issue and happens when a switch is unable to recover to a stable
state. The HASM log contains more details and specific information on this type of failure, such
as one of the following:
- Failover recovery failed: Occurs when failover recovery fails and the CP must reboot.
- Failover when standby CP unready: Occurs when the active CP must fail over, but the
standby CP is not ready to take over mastership.
- Failover when LS trans incomplete: Takes place when a logical switch transaction is
incomplete.
• Software bootup failure
This is an HA bootup-related issue and happens when a switch is unable to load the firmware
to a usable state. The HASM log contains more details and specific information on this type of
failure, such as one of the following:
- System bring up timed out: The CP failed to come up within the time allotted.
- LS configuration timed out and failed: The logical switch configuration failed and timed
out.
After RRD is activated, the admin level permission is required to log in. Enter the supportShow or
supportSave command to collect a limited amount of data to resolve the issue.
Restrictions
The following restrictions are applicable on the RRD feature:
• RRD works only on CFOS-based systems and is not available on AP blades.
• If FIPS mode is enabled, then the RRD feature works in record-only mode.
• RRD relies on the bootprom and the Linux kernel working properly.
• RRD only works during the 30 minutes immediately after the switch boots. If the switch does
not reboot for 30 minutes, then RRD is deactivated.
Collecting limited supportSave output on the Rolling Reboot Detection
1. Log in to the switch on the admin account.
A user account with admin privileges is not able to collect limited supportSave output.
2. After you see the message in the following example, press Enter.