Hot spare or just an extra drive?

[Patrick_3000]

On my primary SCALE server, I have one pool that's a 3x10TB HDD mirror with datasets and another pool that's a 2x4TB SDD pool for VM ZVOLs.

Separately on my network, I have OPNSense installed on bare metal as my firewall/router. I am planning to move OPNsense to a VM on SCALE, which will allow me to get rid of a piece of hardware and reduce energy usage.

However, once I move OPNSense to a VM on SCALE, it will become nearly impossible for me to ever shut down the SCALE server if I need to do so, especially if it happens during the business day since my spouse works from home and needs to be online in order to work, and if OPNSense is virtualized on SCALE, taking down the SCALE server will take down the entire network in my house.

Therefore, I am investigating the idea of adding hot spares to both pools or at least an extra HDD and an extra SDD so that if a drive in either pool ever fails, I can take it offline, add a new drive, and resilver without ever shutting down the SCALE server.

It looks to me like the process of adding hot spares to an existing pool involves adding a VDEV and selecting the spare option. Is this what people recommend? Or is there even an advantage to adding hot spares? It seems to me that I could just install the extra HDD and extra SDD, and if I ever need them due to a drive failure, I could add them to the pool then.

 

[winnielinnie]

If you're going to be on top of resilvers and replacements, you won't really find much benefit of a hot spare. Hot spares make more sense if you want a failed drive to allow another spare drive to "take its place" in the meantime, until you can correct the issue yourself. (Vacation, remote location, unable to be physically present at your NAS server, etc).

 

[Patrick_3000]

If you're going to be on top of resilvers and replacements, you won't really find much benefit of a hot spare. Hot spares make more sense if you want a failed drive to allow another spare drive to "take its place" in the meantime, until you can correct the issue yourself. (Vacation, remote location, unable to be physically present at your NAS server, etc).

Thanks. So if I understand correctly what you're saying, all I really need to do to avoid downtime in my use case is install an extra drive but not add it to a pool. Then, if at some point in the future one of the drives in the pool fails, I can just take it offline from the UI, add the extra drive that's already installed to the pool, and resilver. If so, that makes sense.

 

[winnielinnie]

Whichever suits you the best:

• Have a tested / burned-in drive as a "ready replacement", which you can keep on hand somewhere in the house
• Have a tested / burned-in drive as a "ready replacement", which you can keep within the confines of the chassis (unplugged)
• Have a tested / burned-in drive as a "ready replacement", which you can keep plugged in and powered on, not part of any pool

The only concern I would have is with the last option, since your drive will just incessantly spin before it's ever needed. (Unless you can have the powersave settings keep it in standby, in which you'll have to remember to disable the powersaving settings once it becomes a member of a pool. I'm not even sure how well TrueNAS will prevent the drive from spinning up-down-up-down while its plugged in.)