Query for people out there using vSAN, are people buying there nodes with a hot spare disk in case of a HDD failure in the disk pool?
Am working on a proposal and wondering about if I need a hot spare HDD in the design, I get if a HDD fail then the Disk Group will be inaccessible to VM's and then rely on hardware support to get the disk replaced. Wondering if it makes sense to have a spare disk in the node that can then rebuild the disk pool in event of disk failure quickly to get the node's storage back into action.
Also if there is a HDD failure and a disk is replaced, will there all the objects be copied over to the Disk Pool or is there some magic that means only the missing data is copied. I guess this also depends on the configuration of the disk controller for RAID0 or straight passthrough.
To my knowledge, a HDD failing won't bring an entire disk group offline. The objects will be rebuilt to another HDD, could be on the same disk group or not, assuming there is space to do so. I would want someone else to verify this before taking it as truth. An SSD failing will bring an entire disk group offline as would the controller failing. Thank you, Zach.
If a single magnetic disk in a disk group fails, the disk group will still be available.
The components that were on the failed disk will be rebuilt after some time.
So a long as you have enough disks and disk space, you won't need a hot spare.
Only when a SSD fails, a disk group becomes unavailable.