HI al, intriquing title i know! hopefully enticed you to open this question.
Background. - My client has a single HyperV server using direct storage on DL380 Gen 9 Hardware. Currently 7TB in a Raid 5 Config. They have run out of storage and have asked me to help them move to new storage. they have already purchased 4 14TB Drives. There original idea was to simply back up the existing hyper v guests, blow away the disks and start a fresh on locally installed 14TB. However they didn't realize how much downtime this was going to cause them during the restore. (Another Issue)
They have a second DL380 G8 laying around doing nothing with 64GB ram Dual Xeon(16Cores Total) and 10GIG network. So i thought i would give Truenas a try and install the 14TB disks in that. Both have 10GB links, 2 free so it should work just fine.
Then i discover this 50% rule or more specific degraded performance somewhere after 50% volume utilsation. with 3 14TB disks we are already losing 33% roughly when we create the pool. Then am i right in that to should performance we should never use more than 50% of our already 30% reduced total disk space? Means net gain for purchasing 42TB of storage is 5TB over the existing 7TB they had. Is there any tricks to reduce the impacts of this? or is it just accepted.
How serious is the degradation compared to me installing the disk natively using the Built in raid controller. There current capacity of 7tb as less than 200Gb available and the performance is acceptable. (they are not heavy hitting users) would I expect to see the same on truenas?
I'm thinking just to manage the move using the spare server using local storage now with this limitation
Another question i would like to ask is regarding backup. We use Veeam curently to back up the hyperV guests. With truenas should i use the same approach or back up direct from the trunas server? Will this still enable them to do file level restores or email recovery? Is Veeam suitable for Truenas backup?
Regards for your help
Background. - My client has a single HyperV server using direct storage on DL380 Gen 9 Hardware. Currently 7TB in a Raid 5 Config. They have run out of storage and have asked me to help them move to new storage. they have already purchased 4 14TB Drives. There original idea was to simply back up the existing hyper v guests, blow away the disks and start a fresh on locally installed 14TB. However they didn't realize how much downtime this was going to cause them during the restore. (Another Issue)
They have a second DL380 G8 laying around doing nothing with 64GB ram Dual Xeon(16Cores Total) and 10GIG network. So i thought i would give Truenas a try and install the 14TB disks in that. Both have 10GB links, 2 free so it should work just fine.
Then i discover this 50% rule or more specific degraded performance somewhere after 50% volume utilsation. with 3 14TB disks we are already losing 33% roughly when we create the pool. Then am i right in that to should performance we should never use more than 50% of our already 30% reduced total disk space? Means net gain for purchasing 42TB of storage is 5TB over the existing 7TB they had. Is there any tricks to reduce the impacts of this? or is it just accepted.
How serious is the degradation compared to me installing the disk natively using the Built in raid controller. There current capacity of 7tb as less than 200Gb available and the performance is acceptable. (they are not heavy hitting users) would I expect to see the same on truenas?
I'm thinking just to manage the move using the spare server using local storage now with this limitation
Another question i would like to ask is regarding backup. We use Veeam curently to back up the hyperV guests. With truenas should i use the same approach or back up direct from the trunas server? Will this still enable them to do file level restores or email recovery? Is Veeam suitable for Truenas backup?
Regards for your help