Solved: Hardware problems with a WS-C2950-24

sjhloco · ‎09-20-2010

Hi,

I have been have recently bought some second hand equipment and have been having problems with this one switch. When I start it up after it loads the flash it goes through the tests as shown:

Initializing flashfs...
flashfs[1]: 28 files, 3 directories
flashfs[1]: 0 orphaned files, 0 orphaned directories
flashfs[1]: Total bytes: 7741440
flashfs[1]: Bytes used: 4511232
flashfs[1]: Bytes available: 3230208
flashfs[1]: flashfs fsck took 7 seconds.
flashfs[1]: Initialization complete.
Done initializing flashfs.
POST: System Board Test : Passed
POST: Ethernet Controller Test : Passed
ASIC Initialization Passed

But then this comes up and it goes no further:

ERROR interrupt: PCI Fatal Error on DMA CH0

I tried resetting it and it then starts up showing this and keeps reseting itself over and over again:

=== Flushing messages (00:00:09 UTC Mon Mar 1 1993) ===

Buffered messages:
Queued messages:

Signal 10, Exception code (0x001C)!

Frame 00: SP = 0x80B90E98    PC = 0x8055D560
Frame 01: SP = 0x80B90E98    PC = 0x802C697C
Frame 02: SP = 0x80B90EB0    PC = 0x802C66D4
Frame 03: SP = 0x80B90EF0    PC = 0x802C6910
Frame 04: SP = 0x80B90F60    PC = 0x802C9594
Frame 05: SP = 0x80B90FA8    PC = 0x802CB3D8
Frame 06: SP = 0x80B90FC0    PC = 0x802CAE80
Frame 07: SP = 0x80B90FE0    PC = 0x802CB200
Frame 08: SP = 0x80B91018    PC = 0x802FB7A8
Frame 09: SP = 0x80B91038    PC = 0x801E26A4
Frame 10: SP = 0x80B91050    PC = 0x801E2760
Frame 11: SP = 0x80B91078    PC = 0x8012411C
Frame 12: SP = 0x80B910A0    PC = 0x80124338
Frame 13: SP = 0x80B91178    PC = 0x801C3CFC
Frame 14: SP = 0x80B91190    PC = 0x801C3CE8

=== Crash Fault History ===

C2950 Software (C2950-I6Q4L2-M), Version 12.1(12c)EA1, RELEASE SOFTWARE (fc1)
Compiled Sun 24-Nov-02 23:31 by antonino
Signal = 10, Code = 0x1C, Uptime 00:00:09
r0: 00000000, AT: 00000000, v0: 80B90EC0, v1: 00000001
a0: 80B90EC0, a1: 40000050, a2: 00000004, a3: 80B90EC0
t0: 0000E100, t1: 1000E101, t2: 802F9800, t3: FFFF00FF
t4: 802F97E8, t5: 8072B514, t6: 80B90F88, t7: 00000006
s0: 00000000, s1: 80810000, s2: 00000000, s3: 807567E4
s4: 00000000, s5: 00000003, s6: 80B90F00, s7: 00000000
t8: 00000004, t9: 00000000, k0: 00000000, k1: 00000000
gp: 8078F5A0, sp: 80B90E98, s8: 00000001, ra: 802C697C
EPC: 8055D560, ErrEPC: 00004009, BadVA: 0xF4141BD9, SREG: 1000E103
Cause: 0x0000201C (code 7): Data Bus Error exception

-Traceback= 8055D560 802C697C 802C66D4 802C6910 802C9594 802CB3D8 802CAE80 802CB200 802FB7A8 801E26A4 801E2760 8012411C 80124338 801C3CFC 801C3CE8

Booting after a softreset ...

I have looked on the internet and it seems it could be some type of hardware fault. Someone also suggested trying to use a different IOS, but I cant get in to it to do anything with the IOS.

I was hoping that someone maybe able to explain what any of the error messages mean since I cant seem to find any definate type of answer to where the problem is.

Thanks

Stephen

Bobby Thekkekandam · ‎09-28-2010

Hi Stephen,

In this case the error is occuring during the POST diagnostic on bootup, so unfortunately this is a hardware failure. If the device is covered under a smartnet contract or is still under warranty, I would suggest opening a TAC service request to obtain a replacement.

I don't believe this is the same issue as the bug referenced by Calin, given that the error occurs during operation rather than during POST/bootup. Additionally, the code you are running predates the referenced defect.

Hope that helps,

Bobby

View solution in original post

Calin C. · ‎09-28-2010

Hello!

Do you have any configuration on that switch? I'm asking because I find this:

This problem can affect switches running Cisco IOS Release 12.1(22)EA4. The EtherChannel between two switches flaps, and this message appears:

ERROR interrupt: PCI Fatal Error on DMA CH0

This problem occurs when there is an EtherChannel between two switches, traffic is sent between the switches, and the RSPAN or SPAN source port is configured to monitor sent traffic.

The workaround is to stop the SPAN traffic, and then delete and recreate the VLAN. (CSCeh20389)

http://www.cisco.com/en/US/docs/switches/lan/catalyst2950/software/release/12.1_22_ea12/release/notes/OL16966.html

Maybe you can delete the existing configuration and try then?

Good luck,

Calin

Bobby Thekkekandam · ‎09-28-2010

Hi Stephen,

In this case the error is occuring during the POST diagnostic on bootup, so unfortunately this is a hardware failure. If the device is covered under a smartnet contract or is still under warranty, I would suggest opening a TAC service request to obtain a replacement.

I don't believe this is the same issue as the bug referenced by Calin, given that the error occurs during operation rather than during POST/bootup. Additionally, the code you are running predates the referenced defect.

Hope that helps,

Bobby

sjhloco · ‎09-29-2010

Hi,

I did eventually manage to get into the switch and reset all settings and put on the latest IOS. It did seem to fix the problem temporarliy, but I think this was just a coincidence. I can now get it to work, but each time I boot it up different ports will be disabled as faulty. I found a few more useful things on the internet and they all pointed towards a hardware fault, possibly the ASIC chip.

One thing I found said that each 4 ports where controlled by a separate chip, but I don`t think this can be true since the ports which go down each time are completely random and can be any of the 24. In fact I think it is hardly ever the same ports that go down.

Luckliy the person I bought it off gave me refund and let me keep it, so I cant really complain. As long as you have a bit of patience the switch is usable in a test environment. Its always good to have these problems to learn, I never knew about the Show post command before I started messing with this.

Thanks for the suggestions Bobby and Calin.

Adios