HPE EVA Storage

HP Sizing tool ~vs~ actual disk totals

 
Policia
Occasional Advisor

HP Sizing tool ~vs~ actual disk totals

I have been using the HP Sizing tool to appraise the amount of disks needed to create a certain size disk group. For Example, I have been asked to create three new LUNS on an EVA8100. 1st LUN 1TB, 2nd LUN 2TB, 3rd LUN 1TB. I will need atleat 4TB usable. I used the sizing tool to aggregate the number of disks I needed and came up w/ 22 disks. I am using 300GB 10k's with 95% occupancy. There were no existing disks in this group. The sizer says I need 22 to have a 4.616TB capacity. This is incorrect when I get the disks in the SAN and start carving. I believe I am using the sizing tool correctly. Any ideas what is going awry? thanks ahead of time for the help.
4 REPLIES 4
gstonian
Trusted Contributor

Re: HP Sizing tool ~vs~ actual disk totals

I'm not sure what the sizing tools states but from my calculations I would expect the following from 22 x 300GB Disks with 95% Disk Occ.

H/W - ~5,587.80GB
Raid0 - ~5,308.410GB
Raid5 - ~4,246.728GB
Raid1 - ~2,654.205GB

Are you taking the single disk protection level into account ?
Policia
Occasional Advisor

Re: HP Sizing tool ~vs~ actual disk totals

That is correct. When I created the firt two LUNS of 1TB, and 2TB, I could only had 940Gb left before I hit occupancy. I should have been able to create that last LUN of 1TB right?

How are you calculating the sizes. I am using the "HP StorageWorks sizing tool".

http://h21007.www2.hp.com/portal/site/dspp/menuitem.863c3e4cbcdc3f3515b49c108973a801?ciid=0b083b29d3f021103b29d3f02110275d6e10RCRD

Policia
Occasional Advisor

Re: HP Sizing tool ~vs~ actual disk totals

The disk sizing tool reflects:

300GB 10k/Protection Level 1/95% Occ/
-4560 Usable Cap
-6600 Raw Cap
-4800 RAID5

thanks again for the help.
gstonian
Trusted Contributor

Re: HP Sizing tool ~vs~ actual disk totals

What's available at each stage. i.e. when no disks are created, One disk has been created and then the 2nd. I presume you are creating these as all raid 5 ?

If you can post the Capacity & Occupancy and Available (Raid 0/1/5) at each stage of the process it might explain something.

Also, I presume the disk group isn't levelling at the time (which sometimes causes a disk group to report the wrong available values)