Bidule0hm
Server Electronics Sorcerer
- Joined
- Aug 5, 2013
- Messages
- 3,710
Wow, very detailed and clear post, many thanks ;)
"ZFS always reserves 1/64th of the pool for all the metadata so you can put that into the calculator." it's already in the app (actually it's named checksums overhead but I'll change that to metadata overhead) :D let's call this overhead calculus done :p
So now let's see about the blocks overhead: it's calculable (in percentage or ratio)? And also I thought that the rule of the power of two doesn't apply anymore because of the compression?
The allocation overhead? I didn't think about that. Thanks for the formula, I should be able to implement it pretty easily.
"Also realize that this is just for ashift=12" No problem, the app assumes ashift=12, recordsize=128k and no compression (I should add that in the other thread though) ;)
In case you wonder: "the app" and the "other thread" refers to the RAID size calculator I've made (the link is in my sig).
"ZFS always reserves 1/64th of the pool for all the metadata so you can put that into the calculator." it's already in the app (actually it's named checksums overhead but I'll change that to metadata overhead) :D let's call this overhead calculus done :p
So now let's see about the blocks overhead: it's calculable (in percentage or ratio)? And also I thought that the rule of the power of two doesn't apply anymore because of the compression?
The allocation overhead? I didn't think about that. Thanks for the formula, I should be able to implement it pretty easily.
"Also realize that this is just for ashift=12" No problem, the app assumes ashift=12, recordsize=128k and no compression (I should add that in the other thread though) ;)
In case you wonder: "the app" and the "other thread" refers to the RAID size calculator I've made (the link is in my sig).
Last edited: