ASR9001 sudden reboot after 7 years of uptime

[maddogx]

Hello our A9K1 running 5.3.4, suddenly rebooted this morning, checked environment status was ok, no heating issues, probably no power problems since has 2 power sources.

Please help to identify the reason for reboot.

Found this in logs, could it be the reason?

RP/0/RSP0/CPU0:Mar 14 09:00:08.404 : bgp[1059]: %ROUTING-BGP-5-MAXPFX : No. of IPv4 Unicast prefixes received from x.x.x.x has reached 946104, max 1048576

[smilstea]

Can you check show reboot history?

Thanks,

Sam

[maddogx]

show reboot history location 0/0/CPU0

show reboot history location 0/RSP0/CPU0

both commands output doesnt show today's reboot.

But show-tech-support shows this:

Reload and boot info
----------------------
ASR9001-RP reloaded Mon Mar 14 08:33:29 2022: 3 hours, 36 minutes ago
Active node booted Mon Mar 14 08:33:29 2022: 3 hours, 36 minutes ago

Active node reload "Cause: User Initiated reload"

[smilstea]

That looks like it’s from show redundancy.

Can you check show cli history detail to see if somebody initiated the reload? Or check your syslog server? Syslogs are not persistent across reboots.

It is possible it was a hardware reboot, we can check admin show logging onboard cbc dump-all.

Thanks,

Sam

[maddogx]

show cli didnt show any user reboots.

From the hardware log i see this entry:

Record 58: crc: 0xebcf Sat Oct 29 21:46:40 2016 CBC_OBFL_CPU_RESET_REASON Power Cycle 0x0000 0x0000 0x0000
Record 59: crc: 0x0bb3 Thu Jan 1 02:03:03 1970 CBC_OBFL_CBC_RESET_REASON CBC_RESET_REASON_MCLR 0x0000 0x0000 0x0000

Record 60: crc: 0x49bd Thu Jan 1 02:03:03 1970 CBC_OBFL_BOOTUP_TEMP 21.5 deg C 30.4 deg C 0x0000 0x0000
Record 61: crc: 0xe7bc Mon Mar 14 08:32:18 2022 CBC_OBFL_CPU_RESET_REASON WDOG Power Cycle 0x0000 0x0000 0x0000
Record 62: crc: 0x0bb3 Thu Jan 1 02:03:03 1970 CBC_OBFL_CBC_RESET_REASON CBC_RESET_REASON_MCLR 0x0000 0x0000 0x0000
Record 63: crc: 0x9b95 Thu Jan 1 02:03:03 1970 CBC_OBFL_BOOTUP_TEMP 18.0 deg C 24.6 deg C 0x0000 0x0000
Record 64: crc: 0xffff Sun Feb 7 08:28:15 2106 0x00ff 0xffff 0xffff 0xffff 0xffff
Record 65: crc: 0xffff Sun Feb 7 08:28:15 2106 0x00ff 0xffff 0xffff 0xffff 0xffff

[smilstea]

When the router goes for reboot the time is set in unix/linux to the first available time which is Jan 1 1969, as the router boots the time gets updated as per XR config for clock.

The output you are showing seems to indicate a watchdog power cycle, but it doesn't have any valid data (all 0's). This could still be a legit issue where the watchdog on the cpu timed out, or just a regular reload. Can't say for sure.

I would monitor for further reloads, but if it hasn't happened a second time then its unlikely to be a hardware fault.

Sam