Hi Team,
Imagine that we are using a row store table and during the course of time there is disk failure. Now as we know that row store keeps a WAL for transactions and periodic snapshot which is flushed to disk.
Now as we have a disk failure does the node go down? or it will be still functional.
If you lose one disk on one leaf, and you are running with redundancy 2, and you have at least 2 leaf nodes on 2 different hardware servers, then our standard HA mechanism will failover the the work of the leaf with the bad disk to a leaf or leaves that have replica partitions for the bad leaf.
If you are trying to run everything on one machine, you are out of luck, in less you are using some storage HA technology, like RAID 10.