Featured Article
Clarity 4.0 - New Features and Functionality
by Matthew Zito, Chief Scientist

We typically try to stay away from GridApp-only content in this newsletter, since we figure people get enough sales pitches day in and day out that it's better to set aside the newsletter as a place where we can talk about our ideas around technology, automation, and the overall IT industry. However, this month, I'm going to take a little time to explain why the recently announced version of our database automation software, Clarity 4.0, is a critical step forward for database and application automation, focusing on two of our major new features: model-driven automation and our best practices library.
 
Model-driven Automation

As we met with some of our customers and looked at how they were using the software, we saw that while Clarity was very good at task-based automation - "I need to make a database", "I need to create a cluster" - that even with the automation there were typically multiple steps involved. People wanted to create a cluster, upgrade it, add patches, etc. and had to do each of those steps individually, increasing time and reducing productivity.

The other corollary to this was that the software itself could be smarter about not just doing whatever task that was requested of it, but actually figuring out what needed to be done, and dynamically adapting to meet it. This drove the concept of model-based automation in Clarity 4.0.

A "model" in Clarity represents a desired end state - for example, SQL Server 2005 with SP2 with a hot fix, with certain configuration options set, and some users created. These models can be tied to individual environments and business units, so different models can apply to different environments. When someone needs to build a new environment, they go to the server or cluster, and choose a model. At that point, Clarity will dynamically figure out what about the environment already meets the model - for example, if the user requests a 10.2.0.3 instance based on a certain model, Clarity will examine the server to see if there is already a 10.2.0.3 install that meets the model requirements. If so, then Clarity will prompt to use that ORACLE_HOME, or allow the user to create a new one.

This dynamic automation frees the user from having to know or remember how a server, cluster, or other environment is configured, and allows them to focus simply on what they want the end state to be. In addition, since the model itself contains things like patches, upgrades, etc., the complete lifecycle can be done in one step, with Clarity dynamically figuring out what needs to happen during the automation to make the model complete.

We're the only company today offering this type of no-scripting, out-of-the-box, natively aware automation in the industry, and some of our customers who have been working with the Clarity 4.0 betas agree that it is an exciting next step in automation.

Best Practices Validation

A major piece of feedback we received from our existing user-base is that, while Clarity was quite good at making Oracle and SQL Server easy to install and deploy, often times the server, storage, or other system-level components that the database will use were not set up correctly. The deployment of, say, the Oracle cluster would then fail, and the DBA would have to figure out what was misconfigured. In addition, as they rolled out new types of database environments, the ramp time to figure out the best way to set up their servers was extremely time-consuming.

Based on that, Clarity 4.0 introduces the concept of our validation library. We have gone through the database vendor documentation, knowledge bases, user mailing lists, and worked with our customers to build a library of checks that can validate that everything is correctly configured according to the vendor requirements, or to work around vendor-specific bugs. 

These tests are integrated into the Clarity provisioning process, so that before an activity is performed - say an Oracle installation, followed by an Oracle upgrade, Clarity will run the checks for those activities, present the user with the checks that failed, if any, and then allow the user to proceed. In addition, the checks themselves are audited in Clarity like any other provisioning event, which can be given to security or compliance teams to demonstrate that the server itself was configured according to the vendor best practices.

 So, just how many checks have we built? In contrast to some other database vendor-supplied tools that check 5-10 different configuration options, we have built a library of several hundred checks. These checks range from fundamental items, "Does the user of the database that should be created exist?", to odd bug-related edge cases, "Is the server display resolution set high enough to work around bug X?" We're adding to this content on an ongoing basis, and customers will be updated periodically with new checks, to increase the overall intelligence of the system.

 The value here is that DBAs know that a server is configured correctly without having to check all of those different parameters themselves manually. In addition, it allows the users to always be up-to-date with the vendor best practices, new bugs, and odd edge cases, so if they ever change their configuration to a non-supported or non-standard convention, Clarity will be ready to warn them that the changes made will cause them to hit a particular bug, or generally prevent them from successfully provisioning. 

Conclusion

In the end, while Clarity 4.0 adds some usability and basic enhancements at a core level, it is a huge jump forward for customer environments. The model-driven, out-of-the-box automation, native cluster awareness, and validation library shift the configuration responsibility from the DBA to Clarity, allowing users to focus on the business challenges of tuning and optimizing databases. As Clarity 4.0 ships, we are already preparing for the next version of Clarity, with even more native automation, out-of-the-box support, and greater intelligence around customer environments. I look forward to talking about it with you.

<< back