Anyone care to share some real dedupe/compression savings

2.5:1 on half a dozen systems. But these are all smaller ones.  I spoke with a distributor which also performing the installations and they also see 2.5:1.

Its similar to DataDomain. You need to put a lot of data onto such a system and those system needs to be "large". Putting only 10TB of VM data on a PowerStore you cant expect that only 2.5TB of capacity are needed. Place 100TB on it and you will see a better DRR.

But all the above doesnt matter because Dell guarantee the 4:1 and there are only 4 simply requirements IIRC. I have the document around if needed.

I got 2:1 on a 10TB (physical) system. The compression was off by so much I had to buy more drives.  

First time i see a 4:0 on a small system.... The customer adds around 13TB VMs on a new shiny Powerstore 500T with OS 3.0.1.

Regards,
Joerg

can you share the document or a link to it?   I see 4:1, but not details on how to get the extra drives if you are not hitting those numbers. 

You need to work with your local Dell account team or partner team for the details on the data reduction guarantee and how to claim.

A Customer with 60 VM and a new PowerStore 1200T see 5.7:1  because of having larger MSSQL and Mail(Non-Exchange).

About the 4:1 Gurantee... this paper needs to be signed and send together with your ordering of the HW. Also to be notice that at first Dell have the rights to optimize the system.

Regards
Joerg

Seeing 3:1 with 18TB of VMs. 

For the guarantee - it essential states you can't store already compressed data and you need to store at least a specified amount, which I believe (and could be wrong) is because compression is only used until you go over a used threshold and then dedup kicks in. 

I am just migrating my VMs from our old Pure Storage m20R2 to our new Powerstores 1200T

At the moment there is a mix of Windows and Linux VMs on it (about 270 VMs) and I get 2.5:1 with 16TB used. Our Citrix Servers, SAP Servers and our Windows File Servers are not moved yet. So I think rating will be getting better.

Almost 2 years of use on a 5000T (Upgraded controllers to the 9200T (now running 4.0.1.0) ~6 months ago) Physical storage is 152TB, with a 8.4:1 ratio. I'm not sure if this is a "large" system, but it's running ~250 production VMs, with 2700 Dev. VMs (Dev VMs are all quite similar, quickly deployed from gold images, tested and deleted)

I do feel that we've reached the throughput/resource limit on the 9200T controllers with this workload, as the average CPU use is above 60% during a normal workday. Latency has been an issue for a while now.