Reset Search
 

 

Article

S-Series f/w 7.03.02.0002 is Problematic with Mazama ECC Errors

« Go Back

Information

 
TitleS-Series f/w 7.03.02.0002 is Problematic with Mazama ECC Errors
Symptoms

Upon upgrade the system immediately starts rebooting itself, never becoming fully stable;
  -or-
Upon upgrade the system does boot up, but does not correctly pass traffic.

  • Mazama error messages seen in Syslog and in the output of a 'show support'.
  • Here are distinctive error line examples extracted from one affected system:
  • Message  63/99   Syslog Message            07.03.02.0002   04/28/2010 12:53:54
        <3>mazama[1.interrupt]idx:3-0 0x00e335d0 0x3e87f1e0 0x1101ac10 0x11010000
      
    Message  64/99   Syslog Message            07.03.02.0002   04/28/2010 12:53:54
        <3>mazama[1.interrupt]addr:0x3e87f2c0 0xfefefefe 0xfefefefe 0xfefefefe
     
    Message  65/99   Syslog Message            07.03.02.0002   04/28/2010 12:53:54
        <3>mazama[1.interrupt]Host memory hard ECC error detected.
     
    Message  74/99   Syslog Message            07.03.02.0002   04/28/2010 12:53:51
        <3>mazama[1.interrupt]Host memory soft ECC errors exceeded allowable limits.
     
    Message  84/99   Syslog Message            07.03.02.0002   04/28/2010 12:53:51
        <3>mazama[1.interrupt]Host memory soft ECC error detected.
Environment
  • S-Series
  • Firmware Version 07.03.02.0002
Cause
Version 7.03.02.0002 firmware contains an additional FPGA Hardware upgrade function.
In most cases the upgraded FPGA operates as intended but in combination with certain blade variations can result in the stated symptoms.
It is possible that this will affect some modules in a chassis but not others.
The above error messages are in reference to slot 1, as indicated by the "[1.interrupt]".
Resolution

Upgrade to firmware 7.03.03.0001 or higher.
Release notes state, in the 'Firmware Changes and Enhancements' section:

A programmable hardware (FPGA) upgrade may cause some modules to reset and report several "mazama[1.interrupt]idx:" messages along with "Host memory hard ECC error detected" or "Host memory soft ECC error detected". The upgrade should not be done. Customers experiencing these messages should contact GTAC support.

That release has 7.03.02.0002's FPGA upgrade process removed, with the expectation that a modified version of that process will be reintroduced in a future firmware release.

Workaround:

  • If you are experiencing this issue, please contact the GTAC for assistance.
    Note! This condition is not corrected via a firmware upgrade (e.g. to 7.03.03.0001) or backrev (e.g. to 7.02.03.0001), which updates the firmware version but does not recover the prior FPGA version.
  • If you have successfully upgraded to version 7.03.02.0002 and have not immediately encountered this issue, the only potential consideration is this: Any new blade inserted into the chassis will undergo both FPGA and firmware upgrade, taking the same small yet significant risk of encountering this issue upon upgrade.
  • If you have not yet upgraded to version 7.03.02.0002, it is highly recommended that you not do so, due to the risk of encountering this issue upon upgrade. For this reason the firmware was removed from the web on April 29 2010 (approximately three days after release), and subsequently replaced with version 7.03.03.0001 on May 7 2010.

    In the firmware download page archives, SSERIES-703020002-NOTE.txt ("7.03.02.0002 README") instructs
    Please Contact GTAC for further information regarding 7.03.02.0002
     

Additional notes

Feedback

 

Was this article helpful?


   

Feedback

Please tell us how we can make this article more useful.

Characters Remaining: 255