Guest blog by Yair Horovitz , Chief Functional Architect - Big Data Analytics
Now that HP Discover is over, I’d like to share some of my experiences from the multiple partners and customers that I met.
To set the stage let me just remind you what the value proposition of Operations Analytics is:
Prepare—is about having a Big Data store for IT. Get all your operations data stored and indexed in one place.
Predict—is the ability to infer future behavior from the data you currently have.
Pinpoint—capabilities enable replaying and visualizing all the data to understand when the problem occurred and what caused it.
Even before these meetings I was sure it would be very easy to get people hooked on Pinpoint and Predict. The value of these capabilities is apparent. After all, everyone wants to be able to solve problems faster by quickly finding the root cause of an issue. Obviously no one’s going to object to data that can help them infer whether a problem is about to develop.
I was wondering more about Prepare. Would customers understand the need to store and analyze everything?
Customers have been monitoring systems for years. Tracking metrics and handling events is their daily bread. Why would they ask for anything more?
Well I can tell you that customers ARE asking for more.
The customers I met with clearly understand the need to collect, store and analyze everything.
They know their IT is very dynamic. They know there’s a “shadow IT” effect and that lots of applications only see the light of IT Operations when a problem occurs.
They know there are multiple factors—and some are obscure—that may affect their IT.
That’s why IT customers know they’d like to have all the data available—because they never know when it will come in handy.
Speaking to customers made it obvious to me that preparing all your data for analysis is crucial. They told me over and over again how important it is for them to have all of their data stored and indexed—in an easy to understand format.
In response to these requests, I demonstrated one of the Operations Analytics use cases to show how the Prepare functionality is the answer to their dilemma. In this use case an intermittent application performance problem is diagnosed using metrics and log messages. The problem is identified as a DB locking issue. Obviously the application was updated which caused this issue.
When seeing this, the customer told me it would be great to be able to see additional data, such as planned configuration changes and unplanned upgrades, the type of information that can be found in service management and/or configuration management systems.
Another customer took the “store and analyze EVERYTHING” even further. He said that meteorological data was found to be relevant in diagnosing some his IT challenges. Apparently bad weather changed some of the usage patterns of his subscribers, and caused heavier load on his IT systems. Having this type of data in conjunction with traditional IT metrics would help uncover these correlations, and improve resolution time.
This is why Operations Analytics believes in being prepared - collecting, storing and analyzing everything.
HP Discover gave me the opportunity to display the power of Operations Analytics in person. But it is not the end of the conversation. Feel free to reach out to me in the comments section below if you have any questions, or suggestions regarding data that’s relevant for you.