Hello,
We currently have a hyperconverged cluster with 5x dual E5-2650v2/128GBDDR3 nodes.
We are looking to add start adding a couple of units of E5-2680V4/256GBDDR4 down the road starting with 3 units.
Will these merge into the same cluster? Are there any disadvantages with the mismatched clock speed or uneven RAM allocation? Supposedly the V4s should be more efficient despite having the lower clock speed.
Would you recommend keeping the ceph cluster solely on one architecture (either V2 or V4) while sharing the compute resources across both?
Thanks!
We currently have a hyperconverged cluster with 5x dual E5-2650v2/128GBDDR3 nodes.
We are looking to add start adding a couple of units of E5-2680V4/256GBDDR4 down the road starting with 3 units.
Will these merge into the same cluster? Are there any disadvantages with the mismatched clock speed or uneven RAM allocation? Supposedly the V4s should be more efficient despite having the lower clock speed.
Would you recommend keeping the ceph cluster solely on one architecture (either V2 or V4) while sharing the compute resources across both?
Thanks!