Primary data storage options for the cloud
A comprehensive collection of articles, videos and more, hand-picked by our editors
In part one of this podcast series, Arun Taneja, founder and consulting analyst at Milford, Mass.-based Taneja Group, and SearchCloudStorage.com assistant site editor Rachel Kossman sat down to discuss primary or near-line storage in the cloud. Taneja explained why cloud storage is primarily used for disaster recovery (DR) and backup, and clarified why cloud gateway solutions are essential to near-line storage in the cloud.
In part two, Kossman and Taneja discuss the placement of near-line data in the cloud, starting with a discussion around document collaboration, also referred to as file sharing, and the various cloud solutions that have document collaboration as a built-in feature. Taneja then explains the service levels listeners can realistically expect when placing primary data storage in the cloud, including the types of tier-1 and tier-2 applications that are viable candidates.
Download for later:
- Internet Explorer: Right Click > Save Target As
- Firefox: Right Click > Save Link As
SearchCloudStorage.com: One great example of near-line storage being put in the cloud is document collaboration or file sharing. Arun, can you explain to our listeners why the cloud is such a good option for file and document sharing?
Taneja: Absolutely. Document collaboration is definitely a very good use case for cloud, but I will say straight off that certain clouds have built in technology to allow for collaborative sharing and others don't. The ones that don't can still be used for collaboration purposes, but then that technology has to be built in a gateway. I'll give you examples of both.
In the case of Nirvanix, for example, where they have eight instances of the cloud around the world, they've built their cloud in such a way that the namespace looks like it's one. So if you put a file in the cloud located in California, as an example, and the customer signed up for their data to be kept in the East Coast cloud, in that case when the file is updated in California, that file is automatically updated in the New York cloud. These two files are effectively kept the same, and if someone accesses a file from the New York cloud, and that file is being updated on the California side, the New York person will have to hold until the two files are synced up. So the hard work has already been done by the cloud vendor -- in this example, Nirvanix. That's classic collaboration: Users can ask for a file no matter where they are in the world and they know they're going to get the fresh file. They can do updates from whatever cloud they're interacting with knowing the update will zap itself to all the other locations. That's the technology done on the cloud side.
Listen to part one of the Taneja podcast
In part one of this podcast series, Arun Taneja discusses why IT pros tend to think of cloud storage as only a disaster recovery (DR), archive and backup solution; the technical developments that make storage of primary data in the cloud possible; and why gateways are crucial in placing near-line data in the cloud.
When you don't have that capability -- which is true for a majority of the clouds in the industry right now -- you have to use a gateway, like Panzura. Panzura has created a global file system that effectively takes care of this namespace issue I described. They're doing the heavy lifting and making sure the data that's visible in California is the same [data] that's visible in New York City, Tokyo or wherever that cloud may be.
The customer has to be careful; it's not an automatic or given that just because you have a cloud you have the option for document collaboration. You have to look under the covers to make sure you have the right solution.
SearchCloudStorage.com: So we know from our last discussion that cloud gateways alleviate latency issues, but what kind of service levels should people expect when placing primary or near-line storage in the cloud?
Taneja: A very important consideration [is] what I said earlier -- you can't automatically use a cloud for primary storage purposes; you have to have a gateway, like the one offered by StorSimple. It's important to note that there are a few other examples in the industry for gateways that can make things look like primary storage, but they're few and far between. You have to be careful that these gateways have a full variety of storage applications that come with them, otherwise it won't feel, look and behave like the primary storage the customer is used to in their data center.
More on near-line storage in the cloud
Essential Guide: Adapting hybrid cloud solutions for near-line data storage
Primary data storage made possible by hybrid cloud storage appliances
Using cloud gateways for primary data storage in the cloud
Now what kind of service levels can you expect from that? If it's built in the method I described for StorSimple, then with the service levels you can expect from that -- because of the built-in caching, data deduplication, compression and snapshots -- you should be able to displace practically all tier-2 applications and practically all tier-2 storage. It's very important to understand that if you have a tier-1 application and tier-1 storage today, whether that's VMAX from EMC, VSP from Hitachi, IBM's DS8000 series or Dell's top product line. In those cases, don't think of cloud for primary storage purposes.
But for tier-2 types of applications and storage, which means EMC's VNX series, Hitachi's Adaptable Modular Storage (AMS) line or a whole series of other tier-2 types of storage, the cloud can become a primary storage repository for you, to the extent the service levels that you get from local storage and from cloud will be about the same.