[c-nsp] Sup-720 Spurious / Traceback

[Robert Williams]

Hi All,

Just experienced a load of these on one of our 6500/Sup-7203BXL units:

15:13:01.373 GMT: %ALIGN-3-SPURIOUS: Spurious memory access made at 0x40D42728 reading 0x8
15:13:01.373 GMT: %ALIGN-3-TRACE: -Traceback= 40D42728 40D425B4 40D42618 418D8C74 40306030 40306970 4030A2FC 4313B13C
15:13:01.373 GMT: %ALIGN-3-TRACE: -Traceback= 40D42748 40D425B4 40D42618 418D8C74 40306030 40306970 4030A2FC 4313B13C
15:13:01.373 GMT: %ALIGN-3-TRACE: -Traceback= 40D42754 40D425B4 40D42618 418D8C74 40306030 40306970 4030A2FC 4313B13C
15:13:01.373 GMT: %ALIGN-3-TRACE: -Traceback= 40D42760 40D425B4 40D42618 418D8C74 40306030 40306970 4030A2FC 4313B13C
15:13:01.373 GMT: %ALIGN-3-TRACE: -Traceback= 40D4276C 40D425B4 40D42618 418D8C74 40306030 40306970 4030A2FC 4313B13C
15:13:01.373 GMT: %ALIGN-3-TRACE: -Traceback= 43142C78 40D3867C 418D8C8C 40306030 40306970 4030A2FC 4313B13C 4030B854

Debugging docs say to run this next:

#show alignment
No alignment data has been recorded.

Total Spurious Accesses 2088, Recorded 18

 Address  Count  Traceback
       8    226  0x40D42728 0x40D425B4 0x40D42618 0x418D8C74
                 0x40306030 0x40306970 0x4030A2FC 0x4313B13C
       C    226  0x40D42748 0x40D425B4 0x40D42618 0x418D8C74
                 0x40306030 0x40306970 0x4030A2FC 0x4313B13C
      10    226  0x40D42754 0x40D425B4 0x40D42618 0x418D8C74
                 0x40306030 0x40306970 0x4030A2FC 0x4313B13C
      14    226  0x40D42760 0x40D425B4 0x40D42618 0x418D8C74
                 0x40306030 0x40306970 0x4030A2FC 0x4313B13C
       8    226  0x40D4276C 0x40D425B4 0x40D42618 0x418D8C74
                 0x40306030 0x40306970 0x4030A2FC 0x4313B13C
       8    678  0x43142C78 0x40D3867C 0x418D8C8C 0x40306030
                 0x40306970 0x4030A2FC 0x4313B13C 0x4030B854
       8      9  0x40D42728 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4036EDC0
       C      9  0x40D42748 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4036EDC0
      10      9  0x40D42754 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4036EDC0
      14      9  0x40D42760 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4036EDC0
       8      9  0x40D4276C 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4036EDC0
       8     27  0x43142C78 0x40D3867C 0x418D8C8C 0x403052CC
                 0x403062A0 0x40306970 0x4036EDC0 0x4036F1BC
       8     26  0x40D42728 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4030A2FC
       C     26  0x40D42748 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4030A2FC
      10     26  0x40D42754 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4030A2FC
      14     26  0x40D42760 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4030A2FC
       8     26  0x40D4276C 0x40D425B4 0x40D42618 0x418D8C74
                 0x403052CC 0x403062A0 0x40306970 0x4030A2FC
       8     78  0x43142C78 0x40D3867C 0x418D8C8C 0x403052CC
                 0x403062A0 0x40306970 0x4030A2FC 0x4313B13C

However it has _not_ logged any of these others as traces or spurious accesses?

I’ve just checked all our other chassis and none of them show anything under ‘show alignment’ - i.e. all zero counters.

So we’ve never had this before (we have a lot of these in service) and I’m just curious what thoughts anyone has. Standard documentation says to upgrade IOS and/or log with TAC.

It’s running 12.2(33)SXJ3 so we can upgrade to SXJ4, which was recently marked as a ‘safe’ release. I’m curious though if people think this is likely to be hardware related or a software bug. Again, documentation suggests a software issue, but to have suddenly seen it on one chassis without warning I’m a little sceptical. It’s still performing fine and I’d rather not waste a reload if it’s unlikely to actually fix anything. Only related events were a couple of 3rd party IPv6 BGP sessions reset about 1 minute prior to the most recent events. Not sure if that can be related though, but worth mentioning.

Any advice or pointers appreciated!

Cheers all,

Robert Williams
Backline / Operations Team
Custodian DataCentre
tel: +44 (0)1622 230382
email: Robert at CustodianDC.com
http://www.custodiandc.com/disclaimer.txt