I opened 2 x SSH sessions to each test machine so that I could run 'show proc' while "idle" and while "running commands" to see if there was a difference. Neither switch seems to be maxing out as they are both around 3-4% utilization at "idle" and around 4-5% while "active"
From attached:
- CH1-ZZZZZ is the switch running 16.06.03
- DF1-XXXXX is the switch running 16.06.04
I setup a series of tests running the same command on the same modules on the 2 switches above and timing how long it took from enter to return to command prompt. I chose modules 1-3 (all ports on all 3 modules) and ran the following commands.
- spanning-tree portfast
- no spanning-tree portfast
- description NNNNNN
- switchport voice vlan XX (XX is the site specific VLAN and same as what was currently on the switch to not disrupt any active users).
There was no discernible difference in the time it took for either switch to run these commands. Commands returned to command prompt in ≈ 1:05.00-1:12.00 (time in m:s.ms).
Both sites are relatively new, and we have been making configuration changes at both sites for the past several months. In all previous tests, the CH1 switches performed "quicker" though I don't have exact timers for those. We were also doing larger ranges of modules (including all modules 1-8 except 5 & 6).
Maybe it just 'felt' faster on 16.06.03 than 16.06.04 or perhaps I need to revise my test scenario to "all modules" like we have been doing to see if that is what was making 16.06.03 "faster" ¯\_(ツ)_/¯
