Web and Unmanaged
cancel
Showing results for 
Search instead for 
Did you mean: 

Switch 1910 24P Update issue

 
SOLVED
Go to solution
Highlighted
Occasional Collector

Switch 1910 24P Update issue

Good morning, I have several 1910 24P JE008A switches, I am trying to update them to software version 5.20 release 1519P06 from version 5.20 release 1513P07, so far I have tried 5 of them, 3 successfully but lost 2 of them, someone knows What could be happening? How could I recover those that do not respond? There is no activity on any ethernet port and the console port is not responding.

Thank you in advance.

3 REPLIES 3
Highlighted
HPE Pro

Re: Switch 1910 24P Update issue

Hi @Fx2 !

If the console port does not respond, I am afraid there is no much you can do in this case. First we need to fix this issue. Please, check the console connection - use the same PC, same terminal client software,  same cable (and the same USB-to-Serial adapter if applicable) to connect to working switches just to confirm the issue is not outside failed switches.

The following are the required terminal settings:
•Bits per second—9,600
•Data bits—8
•Parity—None
•Stop bits—1
•Flow control—None
•Emulation—VT100

If you see garbled output, try to change the speed of the port, try 38,400, 115,200 and other values.
If there still will be no output, I am afraid there is no recovery possible and switches can be considered dead.

 

I am an HPE employee

Accept or Kudo

Highlighted
Occasional Collector

Re: Switch 1910 24P Update issue

Hello Ivan B, and thanks for your quick response,

The connection by console is the same for all cases, the thing is that I have more than 100 switches to update to be able to activate Radius and SSH (with the current version it does not allow me), there is a stable version compatible with the current one that allows me do this?

Thank you very much

Highlighted
HPE Pro
Solution

Re: Switch 1910 24P Update issue

1519P06 is the latest from stable versions. To be honest, I have checked our internal s/w repository and do not see anything newer for this model - no betas, no hotpatches... And it is more than a year old, so it must be stable enough for production usage.

I have seen it many times before when perfectly working old equipment breaks on reboot. Just a reboot, no updates, nothing fancy. It is a specific of electronic equipment that some modules that already have defect just die on power cycle. And some 1910 are quite old, especially if we take into account that many switches work in sub-optimal environmental conditions - humidity, temperature etc. etc. Maybe before proceeding with an upgrade you can perform a reboot to verify that the device has potential to survive the upgrade...

Taking all this into account I would not jump into conclusions that the upgrade was the trigger. Even if it was, it could be a defective flash memory that currupted newly uploaded image. But in this case console ports would work and you would see certain error messages. But if after upgrade console ports go down and switches do not light up port LEDs on connection, I would definitely suspect a hardware issue rather that a software one. Another point - if that would be an incompatibility between releases, why 3 of your switches were successfully upgraded and ? Of course, this point makes sense if initial s/w versions before the upgrade were the same on working and non-working devices. 

I am not excluding a possibility of an unkown incompatibility problem that you may hit, just one thing does not fit into this scheme - dead console ports. Normally you should see at least a POST test result on a bootup. And if there is an issue with s/w image, the bootrom should inform you that the image has failed to load. BTW, try the following test - leave the console connected with terminal client open and reboot the switch. Let's see if you get a POST test results at least. 

I am an HPE employee

Accept or Kudo