Chasing a Moving Target: APM in the Cloud - Part 2
Detection, Analysis and Action
February 21, 2013

Albert Mavashev
Nastel Technologies

Share this

In my last blog, I discussed strategies for dealing with the complexities of monitoring performance in the various stacks that make up a cloud implementation. Here, we will look at ways to detect trends, analyze your data, and act on it.

The first requirement for detecting trends in application performance in the cloud is to have good information delivered in a timely manner about each stack as well as the application.
  
We acquire this information via data collectors that harvest all relevant indicators within the same clock tick. For example: response time, GC activity, memory usage, CPU Usage. Doing this within the same clock tick is called serialization. It is of little use to know I have a failed transaction at time X, but only have CPU and memory data from X minus 10 minutes.

Next, we require a history for each metric. This can be maintained in memory for near real-time analysis, but we also need to use slower storage for longer-term views.

Finally, we apply pattern matching to the data. We might scan and match metrics such as “find all applications whose GC is above High Bollinger Band for 2+ samples.” Doing this in memory can enable very fast detection across a large number of indicators.

Here are three steps you can use to detect performance trends

1. Measure the relevant application performance indicators on the business side such as orders filled, failed or missed. And then, measure the ones on the IT side such as  JVM GC activity, memory, I/O rates.

2. Create a base line for each relevant indicator. This could a 1- to60-second sampling for near real-time monitoring. In addition set up a 1-, 10- and 15-minute sample or even daily, weekly or monthly for those longer in duration. You need both.

3. Apply analytics to determine trends and behavior

Keeping it Simple

Applying analytics can be easier than you expect. In fact, the more simple you keep it, the better.

The following three simple analytical techniques can be used in order to detect anomalies:

1. Bollinger Bands – 2 standard deviations off the mean – low and high. The normal is 2 standard deviations from the mean.

2. Percent of Change – This means comparing sample to sample, day to day or week to week, and calculating the percentage of change.

3. Velocity – Essentially this measures how fast indicators are changing. For example, you might be measuring response time and it drops from 10 to 20 seconds over a five-second interval or (20-10)/5 = 2 units/sec. With this technique, we are expecting a certain amount of change; however, when the amount of change is changing at an abnormal rate, we have most likely detected an anomaly.

Now That You Know ... Act On It

After the analysis, the next activity is to take action. This could be alerts, notification or system actions such as restarting processes or even resubmitting orders. Here, we are connecting the dots between IT and the business and alerting the appropriate owners. 

And In Conclusion

Elastic cloud-based applications can’t be monitored effectively using static models, as these models assume constancy. And the one thing constant about these applications is their volatility. In these environments, what was abnormal yesterday might likely be normal today. As a result, what static models indicate may be wrong. 

However, using a methodology comprised of gathering both business and IT metrics, creating automated base lines and applying analytics to them in real time can produce effective results and predict behavior. 

Albert Mavashev is Chief Technology Officer at Nastel Technologies.

Share this

The Latest

May 25, 2023

Developers need a tool that can be portable and vendor agnostic, given the advent of microservices. It may be clear an issue is occurring; what may not be clear is if it's part of a distributed system or the app itself. Enter OpenTelemetry, commonly referred to as OTel, an open-source framework that provides a standardized way of collecting and exporting telemetry data (logs, metrics, and traces) from cloud-native software ...

May 24, 2023

As SLOs grow in popularity their usage is becoming more mature. For example, 82% of respondents intend to increase their use of SLOs, and 96% have mapped SLOs directly to their business operations or already have a plan to, according to The State of Service Level Objectives 2023 from Nobl9 ...

May 23, 2023

Observability has matured beyond its early adopter position and is now foundational for modern enterprises to achieve full visibility into today's complex technology environments, according to The State of Observability 2023, a report released by Splunk in collaboration with Enterprise Strategy Group ...

May 22, 2023

Before network engineers even begin the automation process, they tend to start with preconceived notions that oftentimes, if acted upon, can hinder the process. To prevent that from happening, it's important to identify and dispel a few common misconceptions currently out there and how networking teams can overcome them. So, let's address the three most common network automation myths ...

May 18, 2023

Many IT organizations apply AI/ML and AIOps technology across domains, correlating insights from the various layers of IT infrastructure and operations. However, Enterprise Management Associates (EMA) has observed significant interest in applying these AI technologies narrowly to network management, according to a new research report, titled AI-Driven Networks: Leveling Up Network Management with AI/ML and AIOps ...

May 17, 2023

When it comes to system outages, AIOps solutions with the right foundation can help reduce the blame game so the right teams can spend valuable time restoring the impacted services rather than improving their MTTI score (mean time to innocence). In fact, much of today's innovation around ChatGPT-style algorithms can be used to significantly improve the triage process and user experience ...

May 16, 2023

Gartner identified the top 10 data and analytics (D&A) trends for 2023 that can guide D&A leaders to create new sources of value by anticipating change and transforming extreme uncertainty into new business opportunities ...

May 15, 2023

The only way for companies to stay competitive is to modernize applications, yet there's no denying that bringing apps into the modern era can be challenging ... Let's look at a few ways to modernize applications and consider what new obstacles and opportunities 2023 presents ...

May 11, 2023
Applications can be subjected to high traffic on certain days, which, if not taken into account, can lead to unpredictable outcomes and customer dissatisfaction. These may include slow loading speeds, downtime, and unpredictable outcomes, among others ... Hence, applications must be tested for load thresholds to improve performance. Businesses that ignore load performance testing and fail to continually scale these applications leave themselves open to service outages, customer dissatisfaction, and monetary losses ...
May 10, 2023

As online penetration grows, retailers' profits are shrinking — with the cost of serving customers anytime, anywhere, at any speed not bringing in enough topline growth to best monetize even existing investments in technology, systems, infrastructure, and people, let alone new investments, according to Digital-First Retail: Turning Profit Destruction into Customer and Shareholder Value, a new report from AlixPartners and World Retail Congress ...