Dedicated Failover Hosts, Slot sizes & Performance degradation VMs tolerate...

[T180985]

Currently working with a cluster with admission control enabled & set to Dedicated Failover Hosts. My understanding is this effectively takes these host resources out of the cluster until another host fails.... Using this, are slot still still calculated the same i.e using the largest CPU & Memory reservation? I assume the remaining host resources are split up using these slot sizes? Finally how does "Performance degradation VMs tolerate" affect this if you have it set to 100%? would it essentially ignore the VM resource reservations?

Thanks

[depping · Accepted solution]

No slots are not used at all with this admission control type. you simply set aside the resources of a single host for failure, that is all you are doing. This indeed means the host will not be used during normal operations. If there is no apparent reason to do this I would recommend against it.

I actually described this in the free ebook that published recently: https://pages.rubrik.com/clustering-deep-dive-ebook.html?utm_campaign=authors

https://pages.rubrik.com/clustering-deep-dive-ebook.html?utm_campaign=authors

[T180985]

Thanks! Unfortunately its unclear why its set this way, will need to delve into the design docs again 

Your book is very much on my to read list, thanks for the link