Controllerless Networks

I have a cluster of 14 IAP305's running Version 6.5.1.0-4.3.1.2

Only access points 11-14 are rebooting regulary between 30mins and 1 day 

Spotted these repeated entries in the IAP logs for affected aps every 2 mins. Eventually Reboot reason is alwasy warm reboot.

AP uptime is 51 minutes 51 seconds
Reboot Time and Cause: AP rebooted caused by warm reset

(note: Mac edited, real ip is replaced by 10.xx.xx.xx)

This is only happening on ap 11-14 other aps show no such log entries and have been up for 3 weeks.

Whats is the process that is dieing and restarting, ultimately is that what is casuing the reboot, and why only APs 11-14.

All standard checks have taken place, cabling moving switch ports etc.

Jan 29 12:06:16 nanny[2271]: <303074> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Process /aruba/bin/rngd [pid 4020] died: exited with 0x0
Jan 29 12:06:16 nanny[2271]: <303079> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Restarted process /aruba/bin/rngd, new pid 5435
Jan 29 12:08:07 nanny[2271]: <303074> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Process /aruba/bin/rngd [pid 5435] died: exited with 0x0
Jan 29 12:08:07 nanny[2271]: <303079> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Restarted process /aruba/bin/rngd, new pid 6442
Jan 29 12:09:57 nanny[2271]: <303074> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Process /aruba/bin/rngd [pid 6442] died: exited with 0x0
Jan 29 12:09:57 nanny[2271]: <303079> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Restarted process /aruba/bin/rngd, new pid 7777ith 0x0
Jan 29 12:11:47 nanny[2271]: <303079> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Restarted process /aruba/bin/rngd, new pid 9088
Jan 29 12:13:37 nanny[2271]: <303074> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Process /aruba/bin/rngd [pid 9088] died: exited with 0x0
Jan 29 12:13:37 nanny[2271]: <303079> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Restarted process /aruba/bin/rngd, new pid 10206
Jan 29 12:15:28 nanny[2271]: <303074> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Process /aruba/bin/rngd [pid 10206] died: exited with 0x0
Jan 29 12:15:28 nanny[2271]: <303079> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Restarted process /aruba/bin/rngd, new pid 11541
Jan 29 12:17:18 nanny[2271]: <303074> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Process /aruba/bin/rngd [pid 11541] died: exited with 0x0
Jan 29 12:17:18 nanny[2271]: <303079> <ERRS> |AP AP12 - 24:f2:7f:xx.xx.xx@10.xx.xx.xx nanny| Restarted process /aruba/bin/rngd, new pid 12883

this may be a TPM failure, what's the output of:

show ap tpm-info

What i have spotted is that all the working aps have the following beginning macs
24:f2:7f:c7:96:xx
24:f2:7f:c7:95:xx
tpm output for those is
     Status: Ready

     vendor: ATML
    version: 01.02
        rev: 42.02
       spec: 00.02
     errata: 03

  TIMEOUT_A: 10
  TIMEOUT_B: 10
  TIMEOUT_C: 10
  TIMEOUT_D: 10

All 4 aps that are periodically rebooting all have a different mac series, are they likely to have a hardware variance causing this issue?
Is it a known issue? and is there a software version that resolves it?
This is a busy site so if software upgrade is required wish to move to a resolved version.
24:f2:7f:c7:b3:xx

TPM Output for rebooting APs
TPM not initialized/busy. Please try again later.

the mac addresses should be largely irrelevant, the concerning part is the following:

---

TPM not initialized/busy. Please try again later.

---

Can you try a hard power cycle of the AP(s) just to make sure (support is going to ask you to do that anwyays). If the AP still doesn't come up, then'd I'd get ready to RMA them - you're likely going to be asked to capture the console output of the AP when it's booting (which, if the TPM is bad, there will be some error seen).  That's not always easy to do, and may depend on your ability to replace the APs with spares and take the faulty ones to a location where you can plug them into POE and capture the console.

TL;DR; RMA these APs, pre-empt the requests from TAC by getting console output of one AP that fails and having it ready to send to them, plus the TPM error for all 4 that have problems and that you have tried a power cycle and that other identical APs are OK (etc.)

The Aps have been rebooted / Hard power cycled.

As far as i can tell from the user they are working, they just periodically reboot.

While they are not that easy to access i would not wish to replace then and RMA if the replacement APs do the same.

I have tried looking for identical macs

It cant be a coincidence that the rebooting APs are a different production mac to all the working ones.

My gut feeling is this is likely resolved by software just which one :)

if upgrading is preferred, go for it, there's a slim chance it may solve it, it also may do nothing... let us know what happens either way.