Where, in the AP log, do I see why it was turned off?

[Anonymous]

I downloaded the log of an AP in Virtual SmartZone 3.6.0.222. He was off for 30 minutes on 02/07/2020, starting at 2.15PM. I once saw that it is in the Himem session of the log (I may be mistaken), well, in this session what you have is:

### Himem ###
mem_top = 0xa8000000
High Mem (0xA7FFFC00) size=392
  magic:      0x54545454 0x52434B53
  reset:      0  factory:    0  sw2_reset:      0  sw2_event:    0  post_fail:  0  recovery:   0
  reboot_cnt: 0  total_boot: 20
  reboot_reason: watchdog timeout
  type:       2  index:      5  fis.image: rcks_wlan.bkup
  BootRom:
  GD36 1.0.1.26 - built 16:49:20, Feb 27 2013

  pci reboot: 0
  miniboot:
  magic:      0x6D696E69
  reboot_cnt: 0

  No Bootline

[syamantakomer]

Hi Cesar,

If AP was rebooted, you can check the last reboot reason in support log.

An example output will look like this, from the support log.

### Reboot Reason ###
application reboot

### Reboot Reason In Detail Recently ###
              Reboot Time            UP Time Total Boot Reason
 Wed Jun  3 13:21:48 2020     0 day 00:06:16          3 /usr/sbin/cubic  application, cubic, reboot due to enable mesh
 Wed Jun  3 13:21:48 2020     0 day 00:06:16          3 /usr/sbin/cubic  application, cubic, reboot due to enable mesh
 Wed Jun  3 13:21:49 2020     0 day 00:06:16          3 /usr/sbin/cubic  application, cubic, reboot due to enable mesh
 Wed Jun  3 13:21:49 2020     0 day 00:06:16          3 /usr/sbin/cubic  application, cubic, reboot due to enable mesh
 Wed Jun  3 13:21:49 2020     0 day 00:06:16          3 /usr/sbin/cubic  application, cubic, reboot due to enable mesh
 Mon Jun 22 21:33:44 2020    19 day 07:19:24          1 rsmd_func cubic monitor  AP lost SCG more than 7200 seconds

Himem provides in-depth information about the failure but by default it is not enabled.

Please note that it may not be possible always to find the reboot logs and its relevant information on support logs, because log buffer may get cleared after the reboot.

If you are facing AP reboot issues, I would suggest opening a support case with support logs of impacted AP(s) and support will help you to troubleshoot.

Network connection related common AP reboot reasons:

1- AP cannot reach its gateway.
2- AP lost connection with controller for 7200 seconds- default value or configured value.
3- Power issue.
4- PoE issues on switch.

Less common issues:

5- Kernel panic
6- Watchdog timeout
7- Target failure, etc.

Above three could be related to software bug, memory leak issues and sometimes hardware issues. But only support can confirm if its a known bug/new bug or some other issue, so opening a case with support is must.

Regards,
Syamantak Omer

[Anonymous]

I saw here that many of these AP's that restart and return after a few minutes are model ZF7372, and in the Reboot Reason session, "Watchdog timeout" appears, but in the "### Reboot Reason In Detail Recently ###" session, no date appears corresponding to the time the AP restarted.

### Reboot Reason ###
watchdog timeout

### Reboot Reason In Detail Recently ###
              Reboot Time            UP Time Total Boot Reason
 Tue Mar  3 00:06:36 2020     0 day 00:00:45          4 CLI, user requests reboot via CLI command
 Tue Mar  3 00:07:25 2020     0 day 00:01:34          5 WebUI, rebooting for new software image to take effect
 Tue Mar  3 00:07:25 2020     0 day 00:01:34          6 CLI, user requests reboot via CLI command
 Wed Mar  4 00:37:42 2020     0 day 00:04:38          7 application, wsgclient, reboot due to firmware change
 Wed Mar  4 00:45:05 2020     0 day 00:07:04          8 application, wsgclient, reboot due to country code change
 Wed Mar  4 00:47:21 2020     0 day 00:01:55          9 application, wsgclient, reboot due to ipmode change
 
### Reboot Reason End ###

If the cause is "watchdog timeout", how do you fix it or prevent it?

[syamantakomer]

Hi Ceser,

This is not something you can fix by your own without the help of Ruckus support.

They will analyse the support information, crash signatures, etc and then they will advise how to fix it. They may have to involve engineering team, if it is not a known issue.

As advised in my last comment, you may need to open a case with support, along with AP support logs.


Regards,
Syamantak Omer

[fepam_unipam]

As recommended, I opened Case ID: 01079579