Making the Hard Stuff Easy: IntelliSnap Technology and Hitachi Data Systems (Part 3)

Posted 11/19/2014 by Commvault

Posted in

We started this series looking at how IntelliSnap software works with Hitachi Thin Image technology. Then in Part 2 we looked at ShadowImage. Now we’ll pull it all together into the real-world scenario of using Simpana IntelliSnap® technology to protect and recover applications.

Applications are what it’s really all about. Before we dig into the details, if you’d like a Hitachi perspective on the topic of application recovery you can re-visit a blog from Rich Vining of HDS who provides a good overview.

Let’s get started by outlining the protection process, and then we’ll see how a smart protection process gets you smart recovery. One of the cool parts of IntelliSnap technology is how it pretty much takes care of everything, all the steps required for complete application protection. By applications I mean those big enterprise apps that your organization depends on, like SAP, Oracle, DB2, MS SQL Server and MS Exchange.

First thing we do is provide an application aware agent (or iDA) that frees you up from any scripting or other manual tasks. With the agent in place, we start the process. Follow the numbers in the illustration. Intellisnap technology protection process

Based on policies you set, IntelliSnap software will coordinate between applications and your Hitachi storage and take app-consistent Thin Image snaps (Step 1). The illustration shows two-hour snap intervals, though of course you could configure that however you like. Now, two hours is a decent granularity for a large database. But what if you wanted a tighter recovery point but didn’t want to snap more frequently? IntelliSnap technology also provides managed log backups which can take place on a separate schedule. This is one of those secret sauce features that a lot of people don’t realize exists since they make the assumption that IntelliSnap technology only does snaps. In fact, you can configure frequent log backups to occur — we’re using 30 minute intervals in our example. So now you’ve got a recovery point of 30 minutes instead of two hours. Nice!

You could be done here and have IntelliSnap software clicking away taking snapshots. But you can also go further and allow the software to index your snaps (Step 3) and even back them up to a separate longer-term retention target (Step 4) which can be disk, tape or cloud.

Most daily operational recovery will take place off the snaps, rather than the retention copies, so we’ll focus on that. By way of illustration, let’s say we have an Oracle-based application and there’s a corruption incident at 7:12 p.m. Uh oh! Now you have to get that database back into service as quickly as possible with as little lost data as possible. (As a side note, most of your data loss events will likely be caused by human error or logical corruption. Snapshots are the fastest way to recover from such events.) Hitachi array

Fortunately, since you’re using IntelliSnap technology, you’ve got Oracle-consistent database snapshots every two hours and you’re backing up logs every 30 minutes, all automatically. This means I can run a point-in-time recovery as easily as picking my point in time from a screen that looks like this: Oracle advanced restore options

Just set the date and time you want on the screen and run the restore job. Two things happen when the job starts. First, the bulk of the volume you are restoring will be recovered within the array from the snap image, giving you a very quick recovery, much faster than anything you’d get trying to pull a backup over the wire. Then the IntelliSnap software will replay the Oracle logs back to the point in time you selected. Yup, it replays the logs all by itself. And then it even does any service re-starts needed to bring the app back online.

It’s very important to notice what you don’t have to have on hand to run this recovery. First, you don’t have to have Hitachi storage expertise, which is good news if the disaster happens and you can’t find any of the people who know the ins and outs of the storage. Second, you don’t have to have Oracle expertise, which is also good news if you can’t track down an Oracle expert. This is why we like to talk about “operationalizing recovery.” Without a tool like IntelliSnap technology, every recovery operation is a unique process requiring multiple manual steps and multiple forms of expertise (storage, application, SAN, scripting), all subject to human error at any step of the recovery process. With IntelliSnap technology in place, it’s pick and click, just another routine recovery operation.

This kind of simplicity is even more important when you consider that having a Tier 1 application down and out is a major stress factor in any IT shop. Nerves can get jangled, wires tangled, operations mangled – and then after a few hours of effort, it turns out your recovery didn’t work because during all the stress somebody forgot something and you have to start all over again. And “somebody forgot something” can be as simple as a single character fat-fingered in a restore script.

Bottom line: IntelliSnap technology gives you the intelligence you need to recover your most critical applications in a way that is operationally simple, consistent and reliable. Combine this with the speed and efficiency of Hitachi Data Systems storage and you’ve got a combination that will restore your applications faster than you ever thought possible, and with much less stress!

It’s really very cool stuff. Hitachi and Commvault are helping organizations every day with their toughest data protection and recovery challenges. We can help you too.