1 petabyte Windows ReFS server?

Martin J · July 24, 2018

On 20/7/2018 at 1:00 AM, leadeater said:

Very similar experiences with Nutanix as well, though that differs for the Hybrid nodes/clusters and our All Flash nodes/clusters. Hybrid clusters generally have lower VM performance than when hosted on the Netapps for the same reason you pointed too though the All Flash stuff is blistering fast, the disappointing part though was we never got to compare Nutanix to Netapp All Flash performance only FlashPool SAS. Same issue/advice about keeping node storage and disk counts the same as well.

Interesting. Now I am sort of interested in trying out an all flash NetApp.

On 20/7/2018 at 1:00 AM, leadeater said:

What we've both experienced is what I like to think of as the untold truth about "Web Scale"/"Scale out"/"Hyper-converged"/"HCI" or what ever buzz word of the day someone is using in that while there is some truth in the hype there certainly is hype. If a server is not performing well adding more nodes will not increase the performance at all so never under spec a node, scale up is still just as valid as out.

Yes, when the applications aren't designed for scale out, you might be worse off. for the last 20 years and well, even today, I still see projects where the developers don't design the system for scale out, just scale up. And apparently, database design and usage is still hard, even simple things like using proper queries and indexing of tables, I have to ask that every time they complain about perfomance problems. Well now I am just starting to rant.

On 20/7/2018 at 1:00 AM, leadeater said:

Also for Nutanix which has the CVM VM running on every node which uses ~12vCPU and 48GB ram I'd never buy anything less than 512GB ram per node since so much overhead goes in to running the platform itself.

That's quite bit, no it's gets to be a complicated calculations, we have over 1TB of RAM in each node because it seems that RAM is what we run out of first and because of VMware software licenses, SQL licenses etc and all the usual license costs that we have to factor in.

Well thanks for your input.

Martin J · July 24, 2018

Hello, thanks to everyone for the input. I will be working the next month or so to decide on a solution. I will be adding 45drives to the considerations and have a chat with them. If anyone are interested, I can post our decision here, to tie up the thread nicely, but it might be a few months.

leadeater · July 24, 2018

4 minutes ago, Martin J said:

Interesting. Now I am sort of interested in trying out an all flash NetApp.

The new FAS 8200 controllers we just got have 1TB NVMe cache per head, supports 2TB as well, so should be even faster than what you had before.

5 minutes ago, Martin J said:

And apparently, database design and usage is still hard, even simple things like using proper queries and indexing of tables, I have to ask that every time they complain about perfomance problems. Well now I am just starting to rant.