Recently we've been migrating small distribution sets (~23 switches) from RPVST+ to MST successfully (both with or without spanning-tree loopguard configured in the access). But when we migrate a large distribution set (with around 110 switches) both distribution switches lock-up when we change spanning-tree to mode mst. Our plan was to migrate distribution first, then access (as described in the "config example migration rpvst to mst" on the cisco site). Some specs regarding the network: vlans: ~125; VRRP for redundancy VTP: v3 (used for distribution of the MST instances and VLANS to the access); current spt: RPVST+; virtual ports: ~13.300; vrfs: 12; distribution: 2x 6509; per sw: sup720 3b;msfc3;dfc3b 67xx;4x ws-6748-sfp;2x ws-6704-10g;1x ws-6716-10g;IOS 12.2.33 SXI10; access: 3020 (looped square/triangle combo)/4948 (triangle)/nexus 5596up (looped square/triangle combo); Previously we did sort of the same migration but then from MST to RPVST+ using a script to reload all switches with a modified startup-config (with the new spt mode). We are wondering whether there are other migration plans than one with network outage due to a reload?
... View more