SFE2010P Stack - Needs Rebooting Often

[JamieJ831]

Hello,

I have a stack of 5 brand new Cisco Small Business SFE2010P switches.  These switches are the (new) backbone to my phone network and are in Layer 3 mode containing multiple VLANs and routes.  

Now, they work OK, but every few days or so (and always overnight), they seem to stop working.  All my phones loose connectivity to my PBX (directly attached), I am unable to ping the VLAN interface that links my voice and data networks.  So, I normally get a call around 6am telling me to get my butt into work so I can fix the issue (and I'm normally already on my way!). 

Anyway, rebooting the switch stack fixes the issue.  This isn't always easy - often I can't get into the web admin.  Or i can, but it hangs and hangs, then logs me out (and this is on a server that I have attacked to the stack itself with an IP in the default VLAN subnet). If that's the case, then I'll console into the master and reboot from there.  Usually when I do this, the console just locks up and the switches don't reboot.  So I end up physically pulling the power from the master, and then consoling into the new master and attempting a reboot from there. 

Today, I took a look at my syslog and had the following items:

%2SWDMAIN-F-MSTRTMREXPR: SW2P_dist_tp_change_timer_expiry : TIMER expired on MASTER ENABLE unit   ***** FATAL ERROR *****  Reporting Task: BOXS.   Software Ver sion: 3.0.0.17 (date  08-Nov-2009 time  12:59:16)   0x1842ac 0x181910 0x5d4658 0x425d98 0x428e30 0x746ebc 0x817690 0x817844   ***** END OF FATAL ERROR ** ***

%SYSLOG-F-OSFATAL:     FATAL ERROR: Na05: ABORT DATA exception    ***** FATAL ERROR *****    SW Version  :  3.0.0.17   Version Date:  08-Nov-2009 Version Tim   e:  12:59:16   Instruction            0xA4F928   Exception vector       0x10   Program state register 0xA0000013 0x00a45600 0x00a49e38   0x00757d68 ***** END  OF FATAL ERROR *****

Any suggestions?

Cheer

JJ

[JamieJ831]

Ooops, sorry about the expletive, I didn't think people would find that word offensive.  Better safe than sorry!

[Ivor Diedricks]

JJ,

This sounds like a serious issue. We have not seen this in other customer deployments. I suggest you open a case with the Support Center - they have people and escalation in place to get to the root of the issue and possible solution.

Thanks,

Ivor

[caytos1989]

Did someone find a solution to this problem, I have the same issues...

We have multiple SFE2000, they work normally when they work independtelly, however, when we configured them as STACK, they start working...and the the web browser becomes really slow...and it even hangs up the whole STACK, which means i have to reboot all the switches!!

The only way i can configure the switches is through COnsole port...

this is the error:

%SYSLOG-F-OSFATAL:   FATAL ERROR: GOAH: ABORT DATA exception  ***** FATAL ERROR *****  SW Version  :  3.0.0.17 Version Date:  08-Nov-2009 Version Tim e:  12:59:16 Instruction            0x15FA10 Exception vector       0x10 Program state register 0x20000013 0xffb18541 0x003328d4 ***** END OF FATAL ER ROR *****   

PLEASE HELP!!

PS: I attached an screenshot of the current stack switches

[JamieJ831]

@Carlos

That sounds exactly like the issue I am having. The web admin is very slow, sometimes taking up to 5 minutes to load the homepage after logging in. 

No resolution yet.  I found out that we did not purchase support with the switches.  I'm still going to try to contact support for this, and I have my co-worker who does all the purchasing contact our vendor to see if we can have support added. 

If I find a solution I will post it here for sure. To me this seems like a hardware issue.  I doubt that there is any type of configuration that will fix this.  I'm hoping this is just some bug that could be fixed with a firmware patch, but since the newest firmware is from 2009, I truly doubt there is any firmware updates on the horizon.

[JamieJ831]

Hmmm, from what I can find, this might be an issue with read/writes to an illegal memory location... 

[Ivor Diedricks]

Guys,

Here's one thing we suspect might be the cause. Try turning off Bonjour - instruction is in the attached document.

It's best if you open a case with the Support organization - with the proper escalation in place, your problems will be resolved. Contact information is:

http://www.cisco.com/en/US/support/tsd_cisco_small_business_support_center_contacts.html

Ivor

[JamieJ831]

Thanks - I'll give that a try. Regarding support - I'll be either calling them when I get a chance (things have been a bit crazy this week) or if I can get my support contract going start a ticket online.

[Ivor Diedricks]

When did you buy the switch? Phone Support is free for the first 12 months.

Ivor

[JamieJ831]

Yeah, I have the ability to call support, but it's a matter of a)I don't have time right now to call in b)These switches MUST be up during the work day.  I have 2 call centers that use them

[Ivor Diedricks]

Understood. That makes sense.

[caytos1989]

Guys,

After spending about two to three hours with the client that its having the problems...I;ve gathered all the necesary info to determine the root of the problem!!

First of all, here;s what we got out of the testing we did....

My client's switches (SFE2000) hang up/crashes when they;re working independently or in Stack mode, it makes no difference....

My client uses a PC with Windows Server 2008 R2 and GOOGLE CHROME and INTERNET EXPLORER 8 to configure all his switches, and thats when his Switches hang up...and they stop forwarding traffic, and we loose Web Access to the switch..and the only solution is going through the Console Port (by now, the switches CPU are over 70%!!!) and reload the whole stack...or directly unplugged them from the wall...

So after going through Google Chrome Debug feature...there's an ERROR IN THE PHP CODE that makes the whole STACK go CRAZY, and crashes the whole THING!!! (see attached picture)

The only solution we found , until Cisco finds a better one, is to use Firefox v3.0, or use the console port directly (which you can't do anything from it).!

Also, whenever we did a change, using Firefox 3, the CPU went over 50%...and started going down..and stays at 5% during Peak hours...but just going into the Web Admin page, raises the CPU to over 50% again!

We haven't tried Internet Explorer 9 or Firefox 4.0, but I;m guessing it will crash the whole stack again, because the new browser check the whole code BEFORE running it, to prevent some type of attacking (I guess?)....

I hope Cisco sends us a patch..or makes a Firmware to correct this issue, because everyone will update their browsers sooner or later..and everyone will be screaming!

I'll be waiting for an update!..(btw.. we have tried over 4 SFE2000 switches, and they all crash!)

Thanks (see attached pictures)

PS: Bounjour had been disabled since day one, for security reasons.....so no hope there!

El mensaje fue editado por: Carlos Guardia Prudencio

[David Carr]

Jameson,

Could you please call in to the SBSC at 866-606-1866 and reference this community post.

We need to document a case with you so we can get it escalated.

Thank You

[JamieJ831]

I'll be calling support this afternoon when I can block out some time