HP 2920-48 ProCurve latency (ping)

[MichaeliS]

Hello folks,

we've got some troubles for some time in our environment and need help.

We have four 2920 switches and two 5406r zl2 switches that are configured with rstp. 

For some time we see that when we ping the switches among them, we see often latency with up to 100 ms answer time. I already checked the CPU which does not get higher use than 35 % on the 2920 switches, while the 5406r keep under 10%. The latency is not always visible and appears over and over after some time (approximately after 30 s for further 15 seconds). Before contacting HP support, what can i do to analyse such problems? I'm not really sure if rstp is configured 100% correctly, we just activated it on every switch.  

[Vince-Whirlwind]

Depending on your network topology, you need to choose one switch that will be your Root, and drop its priority down to 0.

Depending on your topology, you might want to lower or raise one or more of your other switches' STP priority as well.

Essentially, you should have a "Core" switch with STP priority 0.
If you have a backup Core, give it STP priority 1.
Switches patched directly to the "Core" can keep the default STP priority.
Switches daisy-chained off a switch patched to the "Core" should have their STP priority raised to 9.
Switches daisy-chained off switches daisy-chained off switches connected to the "Core" should have their STP priority raised to 10.
etc...