And there we have it installed:
View from the back ( ͡° ͜ʖ ͡°):
Very basic just power & the SFF-8088 cable. Currently moving about ~3TB of video over to it. Seeing sustained speeds around 200MB/s. I was seeing speeds as high as 500MB/s but I discovered it was using the servers 256GB of RAM as a write cache. Once it stopped it settled around 200MB/s but that fine. It's plenty fast for archival storage.
Next up, expanding the storage in the backup server (VM server - bottom of the stack) ordered another 10TB drive. Even though RAID0 isn't recommended I can't afford parity right now . I could also use the speed boost. The single backup disk is currently sitting at 72% utilization so I'm going to cut that in 1/2 when the new drive arrives and ~double the read/write performance. I'm not as worried about pool redundancy when the pool itself IS the redundancy. Make sense? I just need to make sure I'm aware the pool is online 100% of the time with no errors. Sounds easy enough right?
-
I sometimes detest using CLI, mostly because my memory is shot since I got my condition... but sometimes just because it's complicated. You know like those "guides" that don't even explain very well what the commands do (if at all), so you're blindly putting commands into it
I almost always prefer GUI, I mean most commands could be done by GUI right? so as long as it doesn't become like a Russian doll situation, and you have multitudes of tabs in the GUI, and nested also within that.
For some it might be quicker on CLI of course, if you know what you're doing, and the needed information to get it done of course... for other like me, it would take far longer and have to go check stuff all the time, LOL
-
I think it's circumstantial. I like having both at my disposal. For some things CLI is faster for others GUI is fine. If you are doing something in a deployment type environment with a lot of machines CLI is far superior as a script will get a machine configured and running far faster than the GUI.
For getting this going from the CLI was pretty strait forward though:
zpool destroy [pool-name] zpool create [pool-name] /dev/sd* /dev/sd*
Beyond that I just re-assign it to the mount-point in the container I use.
It is the fact that mirror/raidz/spare/cache/log/etc are not being specified in the command that it's saying I want RAID0. I'd prefer if it used the word stripe.