On 1st September, Netapp announced a programme to give away $1,000,000 of hardware and services to a non-Netapp customer in order to “prove” their Virtualisation Guarantee – save 50% of your storage in virtualised environments. Now that challenge is available to UK customers – and this time it’s for a cool £1 million.
The concept is pretty simple; Netapp are confident that customers who utilise certain features of their hardware (and software) can use technologies such as snapshot and de-duplication to reduce their overall raw storage utilisation by 50% or greater (using Netapp hardware) or 35% or greater with external hardware on a v-series. The technologies required to be deployed must include:
- Thin Provisioning
- Netapp Snapshot
It’s not hard to see how the savings will be achieved, after all most, if not all, virtual environments will benefit from thin provisioning and the use of snapshots. The use of de-duplication on cloned virtual hosts will improve the savings even further.
There’s no such thing as a free lunch however and the apparent savings do come at the cost of risk and performance. The increase in risk arises because more virtual machines are sharing the same physical storage, so the impact of a failure in a RAID group is more widely spread and more impactful. This explains why RAID-DP is also a requirement in this scenario. Obviously risk can be minimised by not cloning all virtual machines from the same gold master and I’m sure in some environments this may be restricted to test/development setups.
Performance also is likely to be impacted as all of the virtual machines contend to read the same shared blocks of code. With the Netapp architecture, writes aren’t a problem as each write generates a new unique block, however there’s the risk of bottlenecks on reads.
I spoke to John Rollason, EMEA Solutions Marketing with Netapp about the challenge. He emphasised the independent nature of the validation of the results of the challenge, however he also indicated that Netapp would be choosing the participant from the entrants by the closing date of 21st September 2009. This is where things get interesting. What is the correct way to benchmark a virtualised environment? Is there an optimum environment which would benefit he most in this challenge? For instance, if cloning isn’t available with your current storage vendor, should clones be considered fair in the target Netapp solution? I guess only time will tell. Netapp will announce the name of an independent third-party who will benchmark and audit the results. This is expected to be a trusted name in the industry.
Whatever the outcome, it will be interesting to see how other vendors choose to counter Netapp’s challenge. If you’re thinking of entering, you can find the link here.
So, EMC, IBM, HDS, 3Par, do you think you can beat 50%?