PCI Error

[Ceesjan van Hattum]

1 out of 5 boots, i have a pci-error, disabling a few devices. The pci-error of 0/2 causes init-failures on 0/2/4 and 0/2/5 (sam slot interfaces for X25 cards), but not for devices like 0/2/7/1, used for scsi.
My syslog.log also shows that MTBF (mean time between failures) are acceptable. Meanwhile my devices fail and therefor not acceptable for me.

Can anybody tell me what my actual problem is and why it only occures between 10% and 20% of my boots ??


------part of syslog.log:

Apr 8 14:58:22 hostname vmunix: 0/2/5/0 syncio^M
Apr 8 14:58:31 hostname vmunix: 0/3/4/0 syncio^M
Apr 8 14:58:40 hostname vmunix: 0/3/5/0 syncio^M
Apr 8 14:59:10 hostname vmunix: SRA[FTS,c0] (0/2) PCI Error/Abort Condition (gambit = 1)
Apr 8 14:59:10 hostname vmunix: SRA[FTS,c0] (0/2) sam_test = 0x2040, sam_status = 0x10000, pci_error = 0x82060000, iova_error = 0x40002018
Apr 8 14:59:10 hostname vmunix: SRA[FTS,c0] (0/2) Host Access Problem
Apr 8 14:59:10 hostname vmunix: SRA[FTS,c0] (0/2/4) Board Broken; State Reg=0x2
Apr 8 14:59:10 hostname vmunix: SRA[FTS,c0] (0/2/4) fts_handle_disable: old_fault_code = 0, flags = 0x0 code = 0
Apr 8 14:59:10 hostname vmunix: SRA[FTS,c0] (0/2/4) Disabling Device on Maintenace Fault
Apr 8 14:59:10 hostname vmunix: SRA[FTS,c0] (0/2/4) Disabling
Apr 8 14:59:15 hostname vmunix: SRA[FTS,c0] (0/2/4) turning red_light on
Apr 8 14:59:15 hostname vmunix: SRA[FTS,c0] (0/2/4) MTBF Acceptable, (re)enabling
Apr 8 14:59:15 hostname vmunix: SRA[FTS,c0] (0/2/4) Enabling
Apr 8 14:59:24 hostname vmunix: SRA[HRSD] hrsd:1973:hrsdopen: rsd_slot_list = NULL for pci_slot 4^M
Apr 8 14:59:24 hostname vmunix: SRA[HRSD] hrsd:1973:hrsdopen: rsd_slot_list = NULL for pci_slot 5^M
Apr 8 14:59:27 hostname vmunix: SRA[FTS,c0] (0/2/4) turning red_light off
Apr 8 14:59:27 prodob2 vmunix: SRA[FTS,c0] (0/2) PCI Error/Abort Condition (gambit = 1)
Apr 8 14:59:24 hostname vmunix: SRA[HRSD] hrsd:1973:hrsdopen: rsd_slot_list = NULL for pci_slot 4^M
Apr 8 14:59:27 hostname above message repeats 8 times
Apr 8 14:59:27 hostname vmunix: SRA[FTS,c0] (0/2) sam_test = 0x2040, sam_status = 0x10000, pci
_error = 0x82060000, iova_error = 0x4c002018
Apr 8 14:59:24 hostname vmunix: SRA[HRSD] hrsd:1973:hrsdopen: rsd_slot_list = NULL for pci_slot 5^M
Apr 8 14:59:27 hostname above message repeats 4 times
---

[Clemens van Everdingen]

Ceesjan,

What type of machine is this ?
Might be a HW issue or a Firmware issue !

Clemens

[Ceesjan van Hattum]

It is a Stratus DNCP-400-CO.
A so called fault tolerant system.

[Clemens van Everdingen]

Ceesjan,

I believe this is a Lucent box running hp-ux !
Did you contact Lucent about this ?

I could not find anything specific to these boxes.

Clemens.