Currently running OI + NappIT All-in-One on ESXi 5.
Host has Xeon E3-1230, 16GB ECC, M1015 passed through to OI VM, and using 2 x 1TB Mirrored for NFS Datastore for other VMs and 3 x 2TB in RAIDZ for storage. I only get about 3.2TB usable for storage, and as I'm unable to simply add another drive to expand the array I started looking at using deduplication. I've got a Intel 320 Series 40GB SSD I can use for L2ARC if that is going to help. Currently have 8GB RAM allocated to the OI VM. The NFS Datastore is only used by test VMs/not performance critical stuff.
Storage is mostly filled with video files sized from 150MB to 15TB, couple of GB worth of pictures and some music/applications etc.
Am I likely to gain much usable disk space if I enable dedupe, and would my current setup perform OK with it enabled? Would adding the SSD be of benefit?
Also, can dedupe be easily disabled/reversed out of if the performance isn't satisfactory without losing any data/having to start from scratch?
Any other suggestions appreciated.
Host has Xeon E3-1230, 16GB ECC, M1015 passed through to OI VM, and using 2 x 1TB Mirrored for NFS Datastore for other VMs and 3 x 2TB in RAIDZ for storage. I only get about 3.2TB usable for storage, and as I'm unable to simply add another drive to expand the array I started looking at using deduplication. I've got a Intel 320 Series 40GB SSD I can use for L2ARC if that is going to help. Currently have 8GB RAM allocated to the OI VM. The NFS Datastore is only used by test VMs/not performance critical stuff.
Storage is mostly filled with video files sized from 150MB to 15TB, couple of GB worth of pictures and some music/applications etc.
Am I likely to gain much usable disk space if I enable dedupe, and would my current setup perform OK with it enabled? Would adding the SSD be of benefit?
Also, can dedupe be easily disabled/reversed out of if the performance isn't satisfactory without losing any data/having to start from scratch?
Any other suggestions appreciated.