New to unraid and NAS in general. I have a lincstation N1 with two 2TB Samsung Evo Plus m2 drives. I setup one as parity and got the array up and running. It runs a few docker containers but mainly plex. My concern is the array, I came across some articles suggesting that I should use ZFS pools instead because its all flash storage. I have an opportunity now to rebuild it before the storage exceeds some other SSDs I have lying around to back everything up. What is the current thought on this? Should I rebuild it with ZFS pools instead? Unraid version 7.1.2.

Yes, I would recommend a zfs mirror instead of parity + data disk for that setup.

2 hours ago, JorgeB said:

Yes, I would recommend a zfs mirror instead of parity + data disk for that setup.

Just for clarity, you recommend this based on extending the lifespan of the flash storage? Because of the trim issue? If I spend the time to rebuild, I also loose flexibility with expansion in the the future. When I initially scoped out the hardware, I assumed I could expand to 10GB in the future as needed by adding two additional M2 drives and two SSD's. I was under the impression I could add one of the four at a time and just slowly expand. With ZFS, I would have a max of 6TB and have to add them in pairs I believe? That and they would be separate pools of storage where I have to juggle which shares are assigned to which pools? The ZFS example is based on assuming I wouldn't have the ability to back everything up and rebuild it when adding new disks.

Are there people running the standard array with parity disk on all flash without issue? Is it a real concern even with modern flash drives? Some conflicting information around the drives need to rely on their own TRIM processes since Unraid would have trim disabled. I have also read that modern, high quality, flash drives would not be impacted. However nothing definitive. I have read both the lincstation, and unraid documentation and its still not clear to me what the pros/cons are for ZFS versus unraid array with parity on all flash setup.

8 minutes ago, nobler1050 said:

Because of the trim issue?

Mostly becuase of this.

8 minutes ago, nobler1050 said:

With ZFS, I would have a max of 6TB and have to add them in pairs I believe?

With mirrors you could also expand by striping another mirror, if you start with a RAIDZ, 3 devices minimum using the GUI, that can be expanded by one device at a time, CLI only for now.

2 minutes ago, nobler1050 said:

Are there people running the standard array with parity disk on all flash without issue?

Yes, and it's fine, but if the workload is very write-heavy there can be some performance degradation on writes, and if you add more data drives parity will also be more used. Any write to an array data device also means writes to parity, so, e.g., with 5 data devices, parity will have 5x the writes of a normal data device. It may not be a problem for your use case or devices used, but you should be aware

Interesting, maybe buying one more m2 for now, starting ZFS with 3 devices, would improve the expansion in the future. Would ZFS allow me to expand the RAIDZ pool with both 3.5 inch SSDs and the M2 slots? I assume they would all have to be 2TB?

Yes, and it would allow a larger device but only use 2TB