What have I done (Raid configuration question)

[Jediah L.]

I have 4 WD 2TB Reds, I put them in the following configuration:

ZVOL

-raidz1-1 (~3.45TB avail)

-drive1
-drive2
-drive3​

-stripe (~1.84TB avail)

-drive4​

What they heck did this do? :p

I've gathered reading documents that RaidZ1 is similar to Raid5 (with some benefits being software defined) - so I have the availability of the loss of a single disk and still am able to recover (theoretically) - but how does adding the stripe single disk add to or hurt my level of redundancy? What is the controller actually doing with that striped disk?

I did it not having any clue what it was doing adding that last disk; rather because it gave me almost 6TB out of my 8TBs, where-as and I couldn't find any other configuration that gave me more than 4TB or the 8TB.

My goal was to create datasets on the pool of greater and lesser importance and rsync the sets that I felt the raidz1 wasn't good enough for; but then I started thinking... what if the controller puts some of my data on that striped disk instead of the raidz1. What does that mean.. and yes, I am a little confused.

Any insight would be appreciated!

 

[bcorbino]

I unintentionally created the same situation for myself. The short answer is that yes, you and I no longer have redundancy, if that one drive in the stripe (drive4 here) goes bad, ALL your data goes to the Great Bit Bucket In The Sky.

 

[Ericloewe]

http://forums.freenas.org/index.php...ning-vdev-zpool-zil-and-l2arc-for-noobs.7775/

A thorough explanation of the situation.