Research: Top Five Causes of Application Downtime & Delays Revealed

Background

Nimble Storage recently carried out extensive research using data collected by their Predictive Analytics platform Infosight, and analysed more than 12,000 anonymised cases documenting examples of app-data gap related issues. The data was collected from a vast range of IT infrastructures across more than 7,500 customers. Infosight collects between 30 to 70 million sensor data points a day from across the infrastructure in which each Nimble Storage array is deployed. This data provides a comprehensive and granular view into each infrastructure and it should be noted that in 90 percent of cases where issues were detected in these environments, these were remedied by InfoSight before the customer even recognised an issue was present.

Research Findings 

Causes of the App-Data Gap are found across the entire infrastructure stack and contrary to what many IT Managers often assume, are not infact isolated to specific parts of the IT Stack.Whilst this groundbreaking research showed that 46% of issues relating to Application downtime were storage-related, the majority infact (54%) were non-storage related.

Storage related issues comprised of hardware and software issues, software update assistance and occasionally performance issues with examples being failed drives (predictive and proactive replacements) and automated software fault analysis with update recommendations.

The 54% of non-storage related problems could however be identified clearly through the intelligence available from Nimble’s Infosight platform.

Nimble - App Data gap - top 5 causes

In summary – what this research has shown is that if your business is experiencing performance issues and delays in accessing data and applications, the liklihood is that it will be one of the following causes that needs rectifying before you can get back on track and maximise productivity and efficiency across the business:

Configuration issues – 28% : Without Nimble’s predictive analytics portal, the configuration issues that were identified would have been extremely time-consuming and costly to pinpoint and resolve. Every product within an IT infrastructure comes with its own set of recommended best practices which in turn creates a highly complex envrionment where each individual component needs to work together. When this doesn’t occur as is frequently the case – the end-user will be affected by performance delays.

Interoperability issues – 11% : These issues tend to be related to setup configuration with Windows, Exchange and application-level networking. Examples included personnel not following MS-SQL best practices, such as log and database volumes not being separated or MPIO setup on Windows.

Not using best practices (unrelated to storage) – 8% : We found these issues can be related areas such as unaligned IO and networking configuration, including multi-pathing not being setup correctly or incorrect MTU.

Host, compute or VM related issues – 7% : These are issues relating to hosts (Linux, VMs, etc.) as well as setup configuration issues. The challenges encountered included incorrect virtual network configuration, host-side iSCSI setup, UCS setup, and under provisioned hosts.

How to Identify the Top Causes of Application Downtime 

Once your users start to notice reduced performance and the App-Data gap is affecting their working day, it is often a mighty task for IT administrators to locate the root cause of the problem. First instinct is usually to presume the storage environment is at fault and so this prompts the purchase of faster storage – however fast flash alone will not fix ‘non-storage’ related problems. If the underlying issues are not accurately identified, further consequences will occur, impacting seriously on business operations including wasted time, extended downtime, increased user frustration and missed business goals.

To accurately identify and resolve the issues at the point of origination – both quickly and with minimum cost incurred – it is essential that predictive analytics techniques are used which have visibility across the entire infrastructure stack. Furthermore, evaluating potential solutions can no longer be based solely around speeds and feeds or prices – instead legal firms need to validate solutions that incorporate machine learning and predictive analytics which enable:

  • Downtime prediction
  • Downtime automatic prevention
  • Prescriptive resolution
  • Rapid root-cause analysis
  • Cross-stack application of analytics
  • Analytics-driven tech support
  • Measured availability metrics

By introducing data science and machine learning to your IT management processes, you will dramatically close, if not eradicate the App-Data gap, helping your organisation be more productive and therefore raise its competitive position – whilst also freeing up the IT team to focus on high value-add initiatives which will drive growth and profitability.

For a demo on how Nimble’s predictive analytics platform known as Infosight works, or to discuss and potentially resolve any performance issues your business is currently facing, please click below: 

More like this