TFD 12 Primer: Cohesity

Cohesity will be presenting at the upcoming Tech Field Day 12 event, and I opted to do some research on them ahead of time. Cohesity was founded in 2013 and has gone through two rounds of funding, totaling $70 million. Headquartered in Santa Clara, they currently have over 100 employees.

So, what does Cohesity do? They focus on secondary storage, which could include domains such as test/dev, but particularly backup storage / repositories. Cohesity is an interesting mix in that they provide storage, features using said storage, and services, namely backups. In a nutshell, they bill themselves as ‘hyperconverged storage’.

Cohesity’s solution is essentially broken down into three portions:

• DataPlatform – this is the actual storage platform. It currently supports NFS and SMB;
• DataProtect– this is the backup portion of the platform. It is able to integrate with vSphere and leverage VADP in order to do backups;
• Nodes – these are the hardware components. At least three nodes are required to get started, and, according to Cohesity, you can add an unlimited number of nodes. One of the benefits of scaling a single cluster (vs. maxing out at X nodes), is you could achieve better dedupe ratios by having all of your data in one spot.

There are currently two units that Cohesity provides: the C2300 (12 TB raw per node) or the C2500 (24TB raw per node). Each node consists of a hybrid PCI-e flash / spinning disk solution. On the compute side, they use 2 x Xeon E5-2600 per node. To round out the ‘hyperconverged’ feature set, each node comes with 2 x 10 GbE, 2 x 1 GbE, and 1 x IPMI.

In a perfect world, a customer would purchase the right-sized units from Cohesity, rack it up in their environment, and tie it into vSphere. The integration with vSphere step is quite simple in that you just need a hostname (or IP address) for your vCenter, a username, and a password. Once connected, the environment will be presented to the nodes.

Systems that are backed up using Cohesity can also be spun back up in clone-state. For example, if I wanted to run a test against an ERP system, I could spin up a copy from my latest backup. The backup would run on Cohesity’s storage platform, however, I believe that the compute would still run on an ESXi host (selected during the restore process).

From the management side of things, the system runs on an OS dubbed OASIS (Open Architecture for Scalable, Intelligent Storage). OASIS packs a lot of the common features that you’ll see in primary storage arrays (dedupe, compression, replication, QoS, and encryption). One area where this solution shines though is its analytics capability.

If you are able to keep you entire organization’s data in one spot, in the form of backups, there is potentially a lot of information that you could pull from that. For example, instead of having to run a report against all of your servers to look for PCI-related infractions (e.g. stored credit card numbers), what if you could just run it in one spot: on your backups. Cohesity allows this – in fact, you can create custom filters to search the data with.

The management interface itself is built on HTML 5. The system also supports REST API calls, which lends itself to some nice orchestration possibilities (e.g. use a REST API call to build out a new lab environment based on the most recent backups).

Finally, cloud storage is supported as a replication target. Once again, this can be policy driven. For example, I might want to keep everything from the past 7 days on flash, days 8 – 30 on spinning disk, and then days 30 – 60 on AWS. Further to this, you can also spin backups up in some like AWS’ EC2. This is a nice touch. A good use case would situations where you want to spin up a QA lab for nightly automated testing, but the on-premises infrastructure isn’t available.

All in all, I am quite looking forward to seeing what Cohesity is bringing forth for the upcoming Tech Field Day 12.

Disclaimer: I was invited to participate in Tech Field Day as a delegate. All of my expenses, including food, transportation, and lodging are being covered by Gestalt IT. I did not receive any compensation to write this post, nor was I requested to write this post. Anything written above was on my own accord.