I am having issues of my upgrade scripts, stalling for 20 minutes before the reboot and then running a second time. So here is the timing of events.
4:00 - Script starts either as a job or ran manually through script management.
4:11 - swi file is copied from the TFTP server
4:20 - Switch reboots.
4:35 - swi file is copied AGAIN from the TFTP server
4:45 - Switch reboots AGAIN
5:01 - Get the Job Completion Notification stating all Errors (if ran on as job) or when ran manually it will state failure. Basically the reboot command terminates the scrip so it is really a false positive.
Now when I SSH to the switch it self and just copy and paste this script in, all is done within 5 minutes which includes the reboot.
I have created a script that does a pretty good job of upgrading the firmware on our HP/Aruba switches. So far it works on the 2920s, 2910s, 2610s, 2650s, and 2810s. I have not made it to the 3k Series or 5k series of HP Procurve switches as of yet.
Here's the script:
config t
no ip ssh filetransfer
tftp client
exit
write mem
copy tftp flash 192.168.100.100 j9146a/W_15_14_0016.swi primary
y
reload
y
!
Reason for the "no ip ssh filetransfer" and "tftp client" is that not all 200 something switches have their configs totally correct. Something that I will work on with a template later.
So what are your thoughts? Got a better idea?
I read where someone was upgrading the switches using the Firmware Upgrade Operations and change the file .swi extention to .bin and just run it like a Cisco. I have not experimented with that yet.
