[c-nsp] CRS-8-RP V09

[Antonio Soares]

Hello group,

We had 4 faulty CRS-8-RP processors these last weeks. The same problem
happened twice: CRS power-off, CRS power-on and the two RPs went faulty.

There is a pattern.

The SN begins with SAD132900xx and the Hw revision is 09:

PID: CRS-8-RP          , VID: V09, SN: SAD132900xx

Anyone knows if there is a field notice for this ? We opened a case but TAC
said there is no FN but I'm not convinced.

More details about the failures. There are two cases: 

1) the RP hangs just after the memory detection 

Initializing DDR SDRAM...found 4096 MB
Initializing ECC on bank 0
Initializing ECC on bank 1
Initializing ECC on bank 2
Initializing ECC on bank 3
Turning off data cache, using DDR for first time

[hangs]

2) the RP reboots when it is supposed to access the bootflash

Initializing DDR SDRAM...found 4096 MB
Initializing ECC on bank 0
Initializing ECC on bank 1
Initializing ECC on bank 2
Turning off data cache, using DDR for first time

Initializing NVRAM...
Testing a portion of DDR SDRAM ...done
Reading ID EEPROMs ...
..........................Initializing SQUID ...
Initializing PCI ...

PCI0 device[1]: Vendor ID 0x10ee
PCI0 device[1]: Device ID 0x300e
PCI1 device[1]: Device ID 0x1100
PCI1 device[1]: Vendor ID 0x1013
PCI1 device[2]: Device ID 0x680
PCI1 device[2]: Vendor ID 0x1095
PCI1 device[3]: Device ID 0x5618
PCI1 device[3]: Vendor ID 0x14e4
Configuring MPPs ...
Configuring PCMCIA slots ...

System Bootstrap, Version 2.04(20110408:051659) [CRS ROMMON], 
Copyright (c) 1994-2011 by Cisco Systems, Inc.

Acquiring backplane mastership ... successful
Preparing for fan initialization............. ready
Setting fan speed to 4000 RPMs  successful
Reading backplane EEPROM ...
Released backplane mastership ...

Board type is 0x100002 (1048578)

Switch 0 initialized
Enabling watchdog
G4(7457-NonSMP-MV64360 Rev 4) platform with 4096 MB of main memory

..

[reboot]


We made some tests in the lab with one faulty card. This one hanged just
after the Memory tests. After the RAM modules replacement, we got it like
the others: the memory test was ok but the RP rebooted when it was supposed
to load the image.

This was a split boot setup. The disks are FAT32 formatted so the boot
starts on the bootflash and the continues on the disks. I would say the RAM
and/or the bootflash are the faulty components. Unfortunately the bootflash
is not replaceable so we couldn't confirm this suspicion.

The replacement RPs are V10 and V11.


Thanks.

Regards,

Antonio Soares, CCIE #18473 (R&S/SP)
amsoares at netcabo.pt
http://www.ccie18473.net