Highlighted

Cisco blade switches hung

We are operating a number of Dell M1000 blade chassis, each with 2 Cisco ws-cbs3130g-s switch modules. Most are operating fine, but I have one chassis for which the Cisco modules don’t complete their boot. When booting, the STCK light is constant green, and the SYST light flashes for a minute, then both light constant (indefinitely). With either a direct terminal connection or via the CMC command line (connect -m switch-1), there is no activity on the terminal at all (terminal and settings tested on working switches, and all is well).
The curious part — out of 6 chassis, each with two 3130s, I have 4 switch modules that demonstrate the same exact failure mode as described above. I also tried using a spare chassis, loaded with these switch modules (assuming the chassis might be the issue) — same thing.
When viewing these switches via the CMC web interface, they all show as healthy (?), and can be reset via the web interface, but am unable to successfully set dhcp or static IP via the web interface  — no errors are shown, but configs are not changed in the interface.
I find it difficult to believe that all 4 switch modules hard failed the same way. Possibly the firmware is corrupt on these?
Any advice or ideas are appreciated. 
Thanks! 
0 Kudos
4 Replies
Moderator
Moderator

RE: Cisco blade switches hung

Hi,

The LED indicators show what they should show normally. For the system LED blinking green in the POST process and solid green is operating normally. For the stack green means the stack ports are operating. If you are on the console does anything show during post? Did these all fail at the same time?

Thanks,
Josh Craig
Dell EMC Enterprise Support Services
Get support on Twitter @DellCaresPRO
0 Kudos

RE: Cisco blade switches hung

Console (either via direct physical terminal or via CMC console (connect switch-x) shows no activity at all -- ie no terminal output before, during or after boot seems to complete. 

Testing against known working switches modules with the same connection method work fine, full boot cycle shown on the terminal)

Since I "inherited" these servers, uncertain whether they all failed at the same time. 

0 Kudos
Moderator
Moderator

RE: Cisco blade switches hung

In that case there isn’t a way for us to push new firmware onto the switches if we can’t even get to the boot menu to use xmodem.

Thanks,
Josh Craig
Dell EMC Enterprise Support Services
Get support on Twitter @DellCaresPRO
0 Kudos
Detek
1 Copper

Re: Cisco blade switches hung

Hello,

We experienced the same problem running CMC Version 5.22 upgrading to 6.00. Our CBS3130G redundant switches via stack cable became unresponsive.

We had to manually run Firmware update via. RACADM FWUPDATE utility.

The CMC shows the correction version of firmware (6.0). The switches show a status of healthy. However, both switches are now completely unresponsive with a console cable, during POST boot process, and everywhere in between.

All our servers are hard down because the chassis and switch blades are not communicating. This has become an expensive nightmare which has cost us customers!

0 Kudos