Getting incremental value from existing tools such as SCOM GOMEZ and SPLUNK

Introducing the eMite 'SCOM – GOMEZ – SPLUNK' Correlation Extension Creating incremental value from existing tools. So, you have deployed your Gomez APM solution and the value that you are seeing in being able to understand your application performance is fantastic. You’re no longer flying blind and everything is humming along really well, right? Well, not quite.

It’s 9:00am and the application starts running slow. Thank goodness your APM solution has told you, before your users did. But, you have no idea why. So, you call the server team and they take a look at server performance data from SCOM.

It’s now 9:10 and they tell you the servers are running fine. So, you ask them to check databases and nothing’s wrong here either.

It’s now 9:20 and your application is getting slower. You get a call from the server team who use SPLUNK to scour logs and they have now discovered a web service has been intermittently throwing errors. It seems that this is linked to your slow performance.

It’s now 9:35 and your web server is back running by 9:45.

All up, you’ve been down for 45 minutes.

How many of these types of errors do you get every month?

 

emitedashboardeMite's Value

eMite adds a layer of intelligence these above your existing solutions to drastically reduce these types of incidents, providing two key values.

1. Reduce the Mean Time to Repair

Because eMite has already correlated the performance data from your APM tool to the server performance, application logs and transactional databases, when you see slow performance, you will see the cause. This will reduce the time that it takes to determine the problem by quickly pin-pointing root cause.

2. Reduce The Total Number of Incidents

The biggest value is that as eMite trends metrics and predicts when a particular metric will breach its threshold in the future. So, if the volume of transactions in your application increases over time, eMite is going to tell you at what point your application performance is going to impact your revenue. So, rather than reacting to outages, you will be able to avoid them well in advance.

What next?

If you already use eMite, you can ask us for access to the various adapters for these solutions. If you would like to add eMite to your existing applications, contact us for an evaluation.

 

Big Data Analytics for IT

The rise of Big Data is not solely linked to customer loyalty.  Ask an analyst about big data and the first set of examples they are likely to share with you will be about airlines.  The number of tweets from airline passengers, the number of disgruntled passengers, the number of delighted passengers, etc.  The other example is the astounding volume of tweets that get sent when Internet Banking is down.  These are great examples of big data.  Customer loyalty managers should be leveraging big data and in particular analytics on big data to enhance their relationships with their customers.

 

But what about IT?

What about your IT department?

How are they using Big Data?

How are they amassing all the data from all the systems that run your corporation?

How often are they analyzing all that data it? 

How are they predicting what is going to happen and where they should spend their time?

 

We have seen some notable companies start to take a lead here.  Service Providers in particular are waking up to the value of their data.  Some have even starting sharing their analytics with their customers which in turn builds trust.  At eMite, we are in an increasing number of conversations on the topic of avoidance ... around avoiding downtime, avoiding disks filling up, avoiding an increase in service desk tickets, etc.  

 

Smart people in IT are discovering predictive technologies. They are predicting what is likely to go wrong and addressing those issues before they even appear.  Many of eMite's customers are using our Risk Band technology to reduce servce desk tickets with the average reduction being 21% in the first year. 

 

Predictive technology is coming to the IT department.  The big change is in the ease of adoption and ease of use.  You don't need a PhD to get value from the data in your existing systems.  If you want to see the value for yourself then follow this link to request a demonstration. 

 

Reporting in your IT Operations

How have you implemented Reporting within your IT Operations?

 

Do you use:

  • Spreadsheets?
  • Access databases?
  • SQL Server Reporting Services?
  • Dashboard products?
  • An Off the Shelf Reporting Platform?
  • A BI Platform i.e. Tableau, Business Objects, IBM COGNOS, etc?

How frequently can you deliver IT Operations Reports?

  • Hourly?
  • Weekly?
  • Monthly?

What about the data:

  • How intimately do you need to understand the data?
  • Do you rely on specialists that have years of experience with your data and you systems?
  • How many systems do you need to unify?  Do you unify the data?

Navigating your systems:

  • Can you navigate from Service Desk Ticket to underlying cause?  How long does that take?
  • Can you navigate from Service Desk Ticket to the metric? How long does it take you to analyse what went wrong, when it went wrong and why?
  • How long does it take you to integrate a new system (and source of reporting data) into your reporting system?

Connections/Correlation:

  • Does your reporting system connect/correlate metrics to tickets?
  • Does your reporting system correlate Capacity to Availability?
  • Does your reporting system notify you to irregularities … regardless of what they are?

How much time do you spend:

  • Apologising to the business?
  • Tracking down where an incident originated from?
  • In "critical incident" meetings … repeating yourself when each new caller joins the call?
 

Exclusive Pictures of Cantwell's TDF S-WORKS Machine

For those of you wondering what the professionals are riding in the Tour De France we have some exclusive pictures of Jonathan Cantwell's Specialized S-Works machine for you.  Feel free to email This e-mail address is being protected from spambots. You need JavaScript enabled to view it with any specific requests for pictures.

 

cantwells-works-tdf2012

 

cantwells-works-name-tdf2012

 

What's Normal?

Matt Elliott, COO eMite
 

Almost every time we meet a new company in the marketplace we hear statements like "we don't need another monitoring tool, we have already spent alot of money on it (i.e. monitoring)". This is great for us as that means they are ready for Service Intelligence.

The value of Monitoring

There is a vast array of Monitoring tools on the market and they all started by monitoring one silo or another. You can buy network monitoring tools, operating systems monitoring tools, database monitoring tools, synthetic transaction monitoring tools, etc, etc. There is no shortage of tools and no shortage of vendors providing those tools including Microsoft, HP, IBM, BMC, ... to open source providers that are too numerous to mention. There is value in each and every one of these tools. In fact there is also value in implementing many of these tools in the same enterprise as they all focus on a different area and have their heritage in one layer or silo of the IT estate giving them different strengths. Each tool gives you valuable information and many tools overlap, in fact larger vendors have been busy acquiring a broader spectrum of tools so that they can monitor a broader spectrum of IT estate. These tools do a good job of telling what is going on right now in each of those silos. You can get events and alarms and react to those accordingly.

What it doesn't do

Monitoring is valuable in and of itself but it doesn't give you a view into the future, just what is happening now. It also doesn't give you insight into what is normal behaviour. For example if you have a system in a retail bank it could be perfectly normal for that system and its related components to have a peak usage every Thursday at 3 o'clock and a low point on Monday at 10am. If you set a static threshold on that system of say 80% usage and every thursday at 3pm it reaches 90% then you get an "alarm storm" every week. This is both frustrating and annoying and people end up ignoring that time of the week so if there is a real issue at that time it may get you into the press for all the wrong reasons. For these reasons and many more eMite is focused on aggregating metrics, events, relationships, etc from many systems that you already have and performing advanced analysis using an array of out of the box reports and visualizations that you can talior to your preferences.

So what is normal anyway?

Normal behaviour is different in every company. It is not possible to provide an industry standard for normal, nor is it possible to use sources such as SPECint * or other hardware or theoretical models to tell you what your servers, running your applications, using your data on your network could or should be doing. The only way to discover what is normal for your organisation is to aggregate data from many sources and analyze it which is why eMite invented Service Intelligence. It allows you to automate the onboarding of metrics, events, relationships from your monitoring systems, CMDBs, etc you have already implemented regardless of who made them (vendor or bespoke), visually map those metrics, etc to your business services and build your own reports in seconds and your own Dashboards in minutes **. Time is the key which is why we focused on integrating with vendors such as Microsoft, BMC, IBM, HP, Splunk, VMware and many others so that you can focus on analyzing what normal is for you and fine tuning the bad bits of your current normal state using innovations such as eMite Risk Bands which put your normal and your current behaviour on a single screen (with the software doing a whole bunch of smart statistical stuff for you).

Who cares about normal, where is the value?

Understanding your normal patterns of business and what is required of the IT estate to support that now and into the future is important but you may have some more pressing issues we can shed some light on ...

  • Is being able to order just the right amount of storage at the right time valuable?
  • Is removing the burden of monthly reporting valuable?
  • Is decreasing Severity 1,2, 3, etc incidents valuable?
  • Is accurately prioritising IT spend without compromising business services valuable?

    • Enterprise Correlation and Service Models - the missing links

    Watch out for a future blog entries on Enterprise Correlation and Service Models and why they are the missing links between monitoring systems and the business

    * http://en.wikipedia.org/wiki/SPECint
     

    More Articles...

    Page 1 of 2

    << Start < Prev 1 2 Next > End >>