Management Lost through console


Hello Community , I would like to know if some of you have ever experienced a problem when configuring through the console that sometimes it doesn't respond and you must reboot the switch in oder to have again the console CLI working. Do you know a way to fix this behavior in the switch? could it be a compatibility problem with the terminal emulator software that i use (PuTTy)? . I recently have this problem with a customer , the switch model is a summit X670-G2 with 15.6.3.1 firmware and also it happened with a summit X670-G2 with 15.3.3.5 patch 1-9 firmware. SInce they want to use these switches as Core , it's very difficult to explain them why this happened . I Hope that my description is clear enough , perhaps there is some grammar mistakes since english is not my native language , just let me know if you need some details . Kind Regards!

7 replies

Userlevel 4
Marlon,

When facing the issues with console access have we tested the console connection with other switches to confirm the issue isn't with the console cable or PC?

Andrew
Hello Andrew, Yes , i did that test with other switch and the console connection was working fine . I was doing a Lab with the customer so they also used their own console cable and got the same results , at the end we had to reboot the switch. Marlon
Userlevel 4
Hey Marlon,

I could not locate any known issues in those versions of software with the 670G2 platform. Here is a few things to try to narrow down the issue.

1. Try a different emulator such as securecrt, or tera term
2. Change console cable (usb dongle if in use and rj-45cable)
3. Check CPU via "top" command to make sure cpu is normal
4. Check serial settings are correct:

• Baud rate—9600
• Data bits—8
• Stop bit—1
• Parity—None
• Flow control—XON/XOFF

5. connect via telnet or ssh and see if issue follows.
Userlevel 2
Hi Marlon,

We've had the issue a few times with x460-48p switches. Console was working then it stopped for unknown reasons. We tried different laptops thinking it was an issue with them, but that didn't work either. The switch was accessible over the network so we ssh'd into it but console wouldn't work. I couldn't see anything in the logs that gave a reason for the console, though later in the day the stack fully stopped responding and we had to hard boot the entire stack, that reboot then resolved the console problem.

These are the messages we received when the entire stack stopped responding, which is different than you are reporting I know. Not sure if related. Slot 1 was our master and where we were consoling in to. But short and long of it, yes, we have that problem on occasion too, the console locks up.

Thanks

09/01/2015 12:14:44.41 Slot-5: Slot-1 FAILED (1) Conduit recei ve error encountered
09/01/2015 12:14:44.41 Slot-5: System Error 0: Conduit receive error encountered
09/01/2015 12:14:44.48 Slot-4: Slot-1 FAILED (1)
09/01/2015 12:14:44.41 Slot-5: pibConduitMasterRcvOneSlo t(): failed, rc=-6, errno=134, slot=1
09/01/2015 12:14:44.16 Slot-3: Slot-1 FAILED (1)
09/01/2015 12:14:44.18 Slot-2: Slot-1 FAILED (1)
09/01/2015 12:14:44.59 Slot-1: Slot-1 FAILED (1) Error on Slot -1

09/01/2015 12:14:44.59 Slot-1: DM: Lost connection to HAL
09/01/2015 12:14:44.56 Slot-1: Connection lost with pr ocess hal
09/01/2015 12:14:26.58 Slot-1: 2ce2fbfc 0320f809 jalr t9
09/01/2015 12:14:26.58 Slot-1: 2ce2fbf8 00000000 nop
09/01/2015 12:14:26.58 Slot-1: 2ce2fbf4 8f99829c lw t9,-32100(gp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbf0 8fa4002c lw a0,44(sp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbec afa70028 sw a3,40(sp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbe8 sw v0,36(sp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbe4 0000000c syscall 0
Colatuno, Joe wrote:

Hey Marlon,

I could not locate any known issues in those versions of software with the 670G2 platform. Here is a few things to try to narrow down the issue.

1. Try a different emulator such as securecrt, or tera term
2. Change console cable (usb dongle if in use and rj-45cable)
3. Check CPU via "top" command to make sure cpu is normal
4. Check serial settings are correct:

• Baud rate—9600
• Data bits—8
• Stop bit—1
• Parity—None
• Flow control—XON/XOFF

5. connect via telnet or ssh and see if issue follows.

Hi Joe, Thanks for your suggestions , i'll do that for the nex time i face this issue . But do you think that it could be an unknown problem in the XOS software? .
Sarah Seidl wrote:

Hi Marlon,

We've had the issue a few times with x460-48p switches. Console was working then it stopped for unknown reasons. We tried different laptops thinking it was an issue with them, but that didn't work either. The switch was accessible over the network so we ssh'd into it but console wouldn't work. I couldn't see anything in the logs that gave a reason for the console, though later in the day the stack fully stopped responding and we had to hard boot the entire stack, that reboot then resolved the console problem.

These are the messages we received when the entire stack stopped responding, which is different than you are reporting I know. Not sure if related. Slot 1 was our master and where we were consoling in to. But short and long of it, yes, we have that problem on occasion too, the console locks up.

Thanks

09/01/2015 12:14:44.41 Slot-5: Slot-1 FAILED (1) Conduit recei ve error encountered
09/01/2015 12:14:44.41 Slot-5: System Error 0: Conduit receive error encountered
09/01/2015 12:14:44.48 Slot-4: Slot-1 FAILED (1)
09/01/2015 12:14:44.41 Slot-5: pibConduitMasterRcvOneSlo t(): failed, rc=-6, errno=134, slot=1
09/01/2015 12:14:44.16 Slot-3: Slot-1 FAILED (1)
09/01/2015 12:14:44.18 Slot-2: Slot-1 FAILED (1)
09/01/2015 12:14:44.59 Slot-1: Slot-1 FAILED (1) Error on Slot -1

09/01/2015 12:14:44.59 Slot-1: DM: Lost connection to HAL
09/01/2015 12:14:44.56 Slot-1: Connection lost with pr ocess hal
09/01/2015 12:14:26.58 Slot-1: 2ce2fbfc 0320f809 jalr t9
09/01/2015 12:14:26.58 Slot-1: 2ce2fbf8 00000000 nop
09/01/2015 12:14:26.58 Slot-1: 2ce2fbf4 8f99829c lw t9,-32100(gp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbf0 8fa4002c lw a0,44(sp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbec afa70028 sw a3,40(sp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbe8 sw v0,36(sp)
09/01/2015 12:14:26.58 Slot-1: 2ce2fbe4 0000000c syscall 0

Hi Sarah, Thank you for sharing me that experiencie , I had that problem few times too , but with stand-alone switches . Also i got another similar issue before , we lost remote connection to a summit X460-48t that is located in a datacenter , the servers connected to it were working fine , no issues in the network , but we just couldn't have access to the switch remotely , so we had to go the datacenter and check the switch , there we found that problem with the console too that it wasn't responding . At the end we just reboot the switch and it fixed the issue. I opened a case with the GTAC support but there was any log in relation with the event , only the reboot log .
Userlevel 5
How are you connecting to the console port? Are you using a Serial-to-serial, or a USB-to-Serial cable adapter? I have found that USB-to-Serial adapters wont interface properly with the switch if the wrong, or a generic driver is loaded. I had the same issue you're experiencing with Prolific USB-to-RS232 cables while running commands that generate dynamic outputs like top, show ports, etc.

Reply