Wireless Access

 View Only
  • 1.  AP-105s in Reboot Loop

    Posted Jul 22, 2013 04:03 PM

    Deploying 4 new APs to a branch site for a customer (local 650 controller).   The environment is 6.1.2.5

     

    All the APs initially came up and did an image upgrade.  Then upon reboot, they all get in a loop, giving the following error.    I have a case open and am awaiting support, but figured I'd see if anyone had any thoughts.  There was one post on this in the past, but it was deemed a hardware issue in that case.  I am finding it hard to believe that 4 out of 4 APs had hardware problems on delivery.

     

    Testing TPM... TPM driver running in 3204 mode
    i2c_adapter i2c-1: sendbytes: error - bailout.
    i2c_adapter i2c-1: sendbytes: error - bailout.
    Failed -- Rebooting
    killall: nanny: no process killed
    killall: sapd: no process killed
    SIOCSIFADDR: No such device
    Sending SIGKILL to all processes.



  • 2.  RE: AP-105s in Reboot Loop

    Posted Jul 22, 2013 04:54 PM
    Clembo,

    In the console boot process can you check what AOS the APs were coming up on ?

    Do you have other working APs on the same segment ?

    Can you check the show ap license-usage ?

    And on the controller can you check the show log system all ?

    Is cpsec on ?


  • 3.  RE: AP-105s in Reboot Loop

    Posted Jul 22, 2013 05:12 PM


  • 4.  RE: AP-105s in Reboot Loop

    Posted Jul 22, 2013 05:22 PM

    I don't think an RMA will fix this, but a SW upgrade may.

     

    See attached field bulleting about AP-105 TPM issues with specific SW versions.



  • 5.  RE: AP-105s in Reboot Loop
    Best Answer

    Posted Jul 22, 2013 05:25 PM
      |   view attached

    Attachment(s)



  • 6.  RE: AP-105s in Reboot Loop

    Posted Jul 22, 2013 09:58 PM

    Thank you.   I got the same bulletin from a local SE earlier tonight.     I don't recall ever seeing this in any release notes; I wish some of these bulletins were available to partners....would have saved us a lot of time.   Instead we have to RMA all the APs and upgrade code to avoid this in the future.

     

     

    For what it's worth vfabian, there are 100+ AP-105s; but they were manufactured prior the serial numbers in the field bulletin.\



  • 7.  RE: AP-105s in Reboot Loop

    Posted Jul 22, 2013 10:17 PM

    If an AP is in this state already (SAME errors), try upgrading the controller and using this string of commands from the AP local console to factory default the AP.

     

    purgeenv,

    saveenv,

    factory_reset,

    saveenv



  • 8.  RE: AP-105s in Reboot Loop

    Posted Jul 22, 2013 11:09 PM

     


    @clembo wrote:

    Thank you.   I got the same bulletin from a local SE earlier tonight.     I don't recall ever seeing this in any release notes; I wish some of these bulletins were available to partners....would have saved us a lot of time.   Instead we have to RMA all the APs and upgrade code to avoid this in the future.

     

     

    For what it's worth vfabian, there are 100+ AP-105s; but they were manufactured prior the serial numbers in the field bulletin.\


    Thanks Clembo that's good to know...



  • 9.  RE: AP-105s in Reboot Loop

    Posted Jul 23, 2013 12:51 AM

    Typically, field bulletins are not pushed out actively, but they're available to partners and customers on the Aruba support website.

     

    /Onno Harms

    Aruba Networks



  • 10.  RE: AP-105s in Reboot Loop

    Posted Jul 23, 2013 11:37 PM

    Where on the support site would I see this as a partner (or customer)?   I may be missing the obvious, but I can't seem to see it.   Are they just searchable, or are they in a separate listed format?



  • 11.  RE: AP-105s in Reboot Loop

    Posted Jul 24, 2013 02:11 AM

    After logging in on the main support page (support.arubanetworks.com), field bulletins are shown at the bottom under "announcements". The one that I believe covers the issue described here is at the top.

     

    I agree it's not a very user-friendly way to bring this to the attention of our customers (we're looking into ways to improve this), but our support staff should be familiar with the content of these bulletins. It appears that was not the case here; the info is quite recent.

     

    Thanks,

    /Onno Harms