Data services innovations with iguazio

Iguazio (www.iguaz.io), the new player in intensive data centric applications, confirmed their vision and product reality during the first meeting yesterday at Herzliya HQ for The IT Press Tour. I already wrote a few times (June 2016, Sept. 2016 and Nov. 2016) about iguazio and I'm really convinced that the team is making something big that will shake the data processing market landscape.
I work in this industry for almost 25 years, having participated to some stories helped companies to move from insignificant position to key player but this time, it's different, it will be huge.

Orit Nissan-Messing, Co-founder and Chief Architect,
explaining the OSS layer

Cloud has changed everything and Amazon in particular has seriously impacted an industry with a radical new model. It's interesting to see that Amazon was a new comer without any IT role in the past and positions of IBM, HP... were fully shaked. It changes how developers build applications, architect think about IT models, admin deploy applications, vendors sell products and users consume applications but data is still The business asset. Data store is now an abstraction of diverse storage flavors, local or remote, block, file or object, and even with consolidation, the key point is the data virtualization layer and just a few players can do it. Some tentatives have been made but many of them don't scale at all and introduce many limitations. Iguazio has made great developments to mask all this storage complexity with new additional data services.

O. Nissan-Messing pitching the Unified Data Model,
one of the iguazio key differentiator

Also metadata is a classic IT problem, introducing overhead, just remember file system or NFS stack and you get the idea on what I mean by wasted time. And this is even worse with multiple layers on top of each others with normal "classic" integration. Pay attention to that paradox, meta data are essential but also tough to manage and optimize, the majority of today's data management solution use, leverage, enrich and manipulate meta data. Having solutions that understand, optimize and deliver real value around metadata are winning solutions.
It's important to understand that the iguazio solution is not about storage and not about applications but about the integration of the two which make things a different, really different. And by integration I don't mean again a new better simpler platform that leverages classic components and layers. I should say that iguazio is a new animal, something you didn't see in the past, with the mission to deliver the simplest, fastest and lowest cost platform to store and compute data dedicated to new data intensive and centric applications.

Imagine a gigantic data pool, built from a transparent memory/storage hierarchy and controlled by a real-time data catalogue, that offers a common API layer to provide real-time, scalability and unified data access.

Iguazio architecture is super comprehensive able to integrate new data services that will be unveiled in the future by many players, organisations or vendors. You can name this an Unified Data Platform and this is iguazio. In term of speed and performance figures, iguazio beats everything you have in mind:

• Under 100 microseconds,
• 2M ops/sec per node,
• 60x S3 ops/sec vs. others, 100x stream put/sec vs Kinesis per shard, 15x file IOPs vs all Flash NAS, 25x NoSQL ops/sec vs Cassandra or DynamoDB,
• 12GB/s per node,
• 10PB per rack,
• starting below $0.03/GB per month

We had also the pleasure to met Steve Lund, the new VP of Strategic Accounts, based in San Francisco, that will have the mission to boost market adoption in the US.

I invite you to try iguazio, you will love it, no doubt, you will never uninstall it and competition has started to have tough times and nights... Believe me. 2017 will be interesting in that data processing segment.