Once again I’ve been invited by the “Marvelous” Stephen Foskett at another edition of his Tech Field Days, in this case Cloud Field Day 4.
One of the sponsor I was most interested to was Cohesity, that started its presentation on Friday, Aug. the 10th (veeery early in the morning, but welcoming us with an amazing breakfast).
The first leg of the presentation was leaded by Aaron Delp, Director Tech Solutions. He began with a global definition of Cohesity as “Redefining secondary storage: is the first copy of the original data”. So, from this sentence, and from the following too, I understood that, knowing or not, there are multiple copies of data in a datacenter, mostly fragmented, uncomplete, with different purposes, not organized, hard to put together in case of a new purpose to dedicate that copies (so, in this case – another copy of that data!).
The definition of Cohesity as a “DataPlatform” gives me the right idea on how Cohesity compose these data: a platform for all the purposes where all or part of the data are needed. So in few words, the process is opposite: collect data and let the apps and purposes tap into them, instead of let apps and purposes collect data individually tapping into lot of replicated data.
It goes withut saying that API availability is nomore a plus today – better, it’s a minus if you don’t offer them, of course Cohesity allows them.
Now, the interesting part related to the cloud (isn’t it CFD after all?): the virtual edition of the product that leverages the on-prem appliance running locally on vSphere, connected to a Cloud edition appliance. A strong architecture defines the main datacenter hosting a physical Cohesity appliance linked to branch offices (or apps) where a virtual Cohesity appliance will run in sinergy with the main one and, at the end of the chain, a Cloud edition of the app replicating the whole bunch of data to a public cloud.
To wrap up this first part of my blog series dedicated to Cohesity, i’d like to pick up the key cloud differentiators that Aaron mentioned in his session:
- Global Space efficiency – in a cloud environment this doesn’t mean only storage efficiency, but also network (ingress and egress aren’t for free), dedup before transit to be efficient across the wire.
- Hybrid & Multicloud Mobility – no lock-in for cloud provider choice, nor for on-prem or in the cloud, but move wherever and whenever you want or need.
- Policy Driven automation – simply move and manage data according to templates.
- Cloud/Apps Integration – architecturally, adapters to protect natively applications in the cloud (a.e. protect Exchange, MSSQL, Oracle on public cloud) or integration of primary storage with cloud storage.
I will soon follow this post with a deep dive in cloud as an extension of on-premises data centers showing a couple of use cases, both presented at CFD4.