winnielinnie
MVP
- Joined
- Oct 22, 2019
- Messages
- 3,641
Currently, my boot-pool lives on a single 64GB SSD. I have also relocated my System Dataset, logs, and reports to this boot-pool.
I understand the risks of logging and reporting to an SSD: these drives have limited write cycles. What I have also read is that this risk is usually over-exaggerated, considering that it takes a LOT (as in a LOT, LOT, LOT) of data being constantly written to kill an SSD, and that the drive's controller already spreads out the writes over its entire capacity.
My question is more specific to FreeNAS. Will a boot drive failure (or system dataset) failure on a live system negatively effect or corrupt the user's storage pools?
As it stands now, is there anything outside of a ZFS storage pool's metadata that relies on anything living inside the system dataset or boot-pool? If, for some reason, I suddenly yank out my boot drive (essentially simulating sudden failure of my system dataset and boot-pool during a live running system), will it cause corruption to my storage pools? Does the risk of irreversible corruption increase if my storage pools were in the middle of reads/writes through an SMB share? Is this any different than ZFS resiliency through an electrical power outage?
I understand the risks of logging and reporting to an SSD: these drives have limited write cycles. What I have also read is that this risk is usually over-exaggerated, considering that it takes a LOT (as in a LOT, LOT, LOT) of data being constantly written to kill an SSD, and that the drive's controller already spreads out the writes over its entire capacity.
My question is more specific to FreeNAS. Will a boot drive failure (or system dataset) failure on a live system negatively effect or corrupt the user's storage pools?
As it stands now, is there anything outside of a ZFS storage pool's metadata that relies on anything living inside the system dataset or boot-pool? If, for some reason, I suddenly yank out my boot drive (essentially simulating sudden failure of my system dataset and boot-pool during a live running system), will it cause corruption to my storage pools? Does the risk of irreversible corruption increase if my storage pools were in the middle of reads/writes through an SMB share? Is this any different than ZFS resiliency through an electrical power outage?