During the recent webinar, HP IT: Getting to Root Cause Faster with Data Analytics, Gary Brandt from HP IT along with Sancha Norris, HP Product Marketing Manager, shared three use cases of HP Operations Analytics, where his team achieved results in minutes rather than days or weeks.
There were a lot of good questions at the end of the webinar, and I didn’t want you to miss out on this vital information.
Sancha: So our first question: Gary you mentioned in your network example how it took 30 minutes to apply out to the analytics to the problem. Can you comment a little bit about how hard it was to create the dashboard and what kind of training is required to use Operations Analytics?
Gary: We don’t have any hardcore data scientists on our team to help build tools. We implemented Operation Analytics with dedication so we could hit the ground running. Out-of-the-box there are lot of different dashboards that come with the product. We take them and tailor them to your specific needs and use them like a query like language called AQL or Analytic Query Language. Basically, you can take that from the out-of-the-box dashboard and use that as a starting point to tailor your own or make changes or customizations to it. But all of the visual pieces make up a little pallet that you can pick from and show how you want to represent the data.
It’s been used from a training standpoint, which is a relatively low learning curve. You need to have some kind of understanding of the SQL structure – SQL language structure. But, if you understand that and understand the data that you’re dealing with, it was network data or something then you can create your dashboards pretty easily.
Sancha: The second question goes to your exchange pilot example. Will it include other processes? That is, network and virtualization in your ecosystem?
Gary: It already does actually. I know we went through a lot of material really fast, but on that exchange dashboard we created, we were pulling network performance data, and Exchange performance data as well as syslog data. One of the things that I mentioned is that we wanted to leverage our existing investment. From a monitoring logging perspective, we already collect syslog data on all of our network devices into ArcSight Logger and also we use NNMi (Network Node Manager i) and NPF (Network Performance Server) that monitors and collects metrics for the purpose of monitoring.
For that exchange dashboard we tapped into those products I just mentioned. Operation Analytics has native integrations into them, so we were able to quickly pull that information into Vertica, and then once it is in Vertica, it is fair game for any analytics and dashboarding that you want to do. This is why we included the network side.
On the virtualization piece, if you’re taking about VMs, we do have our cloud (as I mentioned), all the data from the VMs on the system’s level, CPU Memory Disk, resource pools and things like that, we’re already pulling all that information into our instance of Operations Analytics today.
Sancha: The next question is one I might take since it’s a product question. Do you need to have other HP Management Products to use Operations Analytics; can I bring in my own data?
The answer to that is yes. The value of Operations Analytics is really that you can bring in whatever data source that is in your environment.
As a matter of fact we have close to about 50 data collectors today, of those only about 20 percent are from HP, and are BSM related. Operations Analytics can be standalone and used with whatever other tools that you have in your environment.
We also have the access to something called a flex connector that can create custom connectors if you have something very specific. The idea is for you to bring as much data as possible into Operations Analytics and use that to get insight.
Sancha: Our last question in the pipeline for today, how is the data in the dashboard that you showed in your presentation related to each other, can you explain that a little bit?
Gary: Because we are pulling in data from different sources, it’s not always apparent that the things are related. Within Operations Analytics, we offer the ability to define a topology where you can say that two or more pieces of information have a relationship. We were able to do that and that’s actually how we correlated a lot of the data that we display in our dashboards with each other. Also, as part of the collection process, you can tag data if you want to give it a category or a unique label. Then when it using its query language, it can match those. Anything that's been tagged with a similar value can be pulled together in a particular view.
The third thing that we are doing is we use UCMDB (Universal Configuration Management System) which keeps track of topology in our environment. It keeps track of applications, servers and services that are all connected together. We can pull some of that information into Operations Analytics through BSM today which gets that information from UCMDB, which is another way we can relate to data.
From those three tools and mechanisms we are able to write queries and can join information together in those analytic dashboards.
GaryBrandtHP Global IT Functional Architect, Operations Service Enablement
Gary has over 20 years of experience in the IT industry focusing on enterprise-scale business solutions and operations. He is the principle architect for designing and instrumenting monitoring and operational analytic solutions for HP’s private cloud and traditional data centers. Gary is a subject matter expert with many HP Software products and is using HP Operations Analytics to enable HP IT operations teams to analyze trends and patterns in data to isolate, solve, and proactively prevent production problems.
Sancha Norris Sr. Product Marketing Manager, Operations Analytics, HP Business Services Management
Sancha leads the product marketing responsibilities for Operations Analytics at HP, the first HP analytics solution built on the HAVEn big data platform. Previously, she was the solution marketing manager at CyberSource (a Visa, Inc. subsidiary) where data analytics was the cornerstone of its fraud management solution. Sancha has experience in business intelligence since the days of traditional BI and database to now big data analytics.
HP Operations Analytics is a Big Data analytics solution that helps IT use all the insights hidden in system silos to resolve root cause of failures faster and improve future operational performances.