Procurve 5308xl and MSTP/RSTP Configuration (811 Views)
Reply
Occasional Collector
BlueCrescent
Posts: 1
Registered: ‎10-24-2011
Message 1 of 1 (811 Views)

Procurve 5308xl and MSTP/RSTP Configuration

Hello All,

 

For some time now, we have been receiving complaints about network performance on our 5308xl switches.  After sniffing the packets, performing tests with STP and topology updates, I started to review the configurations of the 4 switches (one in a different geographical location).  At the sites where it works, I noticed a difference to how we setup our vlans.  I went further into configuration concepts/standards for STP.

 

I am using version E.11.10, ROM E.05.05 and it's default STP operation is MSTP.  I had found a very good document on the net talking about MSTP configuration concepts for the switches and had some more specific questions regarding it.  So any advice and/or help would be appreciated.

 

Configuration:

 

We use a vlan standard according to the location.  For example, Breda Netherlands, we use a 3 for location of this site.  This three is used for the ip segment (x.3.x.x) and for the vlan, 3.  Vlan 3 is now the primary vlan.  When we initally configured STP, it was under the default vlan configuration of the switches, vlan 1, which was the default vlan.  Now, vlan 1 is still present, yet no ports are configured under it (no tagging...).

 

The MSTP instance id is the default as we did not specifically configure an id number.  Therefore it is "0".  STP was configured first before the new vlan, vlan 3, was created.  Internal path costs for each port is either auto or 200000.  The forced version on this site is RSTP.

 

ports attached to end devices are configured with bpdu filtering and protection on them.  They also have lacp disabled.

 

LACP is statically configured for our server at that location for NIC teaming.

 

We do have 6 modules on the switch where module a is a gb module.

 

Issue

 

Looking at the debug-counter for stp, I see topology changes, which I now do not understand.  This site also seems to be experiencing a lot of network performance issues.  Specifically with topology changes.  I do not understand why there are so many topology changes for this site.  The only auto-edge/non-edge ports are the uplinks to the router and/or our firewalls.  This is at every site.

 

At another site the vlan is not company standard and therefore is in the default vlan and this site is not experiencing any network performance issues.

 

What could it be

 

Could it be that STP needs to be reconfigured as they new vlan, vlan 3, was created?

Could it be that the default instance, instance 0, is not connected to the new vlan 3, which is the primary vlan

Could it be how LACP is configured, static trunk?

 

Any advice is greatly appreciated.

 

TIA

blue

 

 

Please use plain text.
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation