I'm looking at a few hyper converged servers, and two Catalyst switches. This is a new clean slate. My concern is not with the hardware, but rather the best way to configure the switches for redundancy to keep the hyper converged systems online (think disk replication etc) in case of switch failure or maintenance.
I have VSS on another switch pair that works well. I get to use all uplinks, and can attempt to ISSU if really wanted to gamble. Having one data plane does bother me.
I'm leaning towards HSRPv2 or GLBP. I could double up the links (2 to each switch / device) to get the same available throughput as VSS if I really had to.
Any thoughts or suggestions?
No comments:
Post a Comment