PerfLink: February 2019

Mobile Device Performance Testing

Due to high usage of the internet and ease of doing the e-operations like transferring the money, ordering the food, playing the internet game, listening to the songs etc. all these daily needs are increasing the demand of the mobile. Now, a single mobile can do multiple things, not only on the internet but offline also like capturing the high definition images, video and audio recordings, storing and managing the contacts etc.

Just like a server now it is mandatory to do a performance check of the mobile device so that it can perform well with the provided resources (hardware) and give a better user experience. But mobile device testing is not as similar to the server or application performance testing.

Neotys has published an informative white-paper on Mobile device performance testing and challenges. If you are involved in such kind of performance testing then this white paper will be a treasure for you. Refer to the white-paper link below:

Tips on Performance Engineering

Following the 2019 trend in the performance testing world, skill upgrade must be required for a native performance tester. When we read any advanced level topic related to performance testing then two important terms always come across. Now, the question here is how to implement these two terms in the project so that a performance tester can switch from a native process to the advance approach?

An article is published by JM in focusaps which bridge the gaps in performance engineering. In simple words, he gives the tips on performance engineering in DevOps term. Read the article and grab the knowledge on how small things make big differences. Also, update your skills as per market demand:

Reasons for High CPU Usage

High CPU is one of the common bottlenecks found during performance testing. High utilization of CPU may occur in any type of performance test. Here, I am not referring to the low or high timely CPU spikes which come rarely. I am pointing to the high CPU utilization which lasts till the test ends or sometimes even after the test completion. There could be multiple reasons which make such type of situation.

I found an informative article which describes the possible and most common causes of high CPU utilization. I would suggest to save these causes in your memory because in future when you try to resolve the bottleneck related to high CPU utilization then this learning could be a key treasure for you.

Performance Engineering with DevOps Methodology

Performance Engineering is growing day by day. IT companies whether they are service-based or product based, are demanding for the resource who has knowledge of both performance testing as well as performance engineering. As we know that performance testing is based on the black box concept where performance engineering forces us to understand the architecture of the application and deep-dive into backend functionality of the system. So, what could be your strategy to evolve performance engineering in your project with DevOps methodology. Understand by this post (a high level summarized article):

JMeter - Learning Material

JMeter is an open-source tool and nowadays it is in high demand. Day-by-day new plug-ins extend the features of JMeter and make it a stronger competitor for the existing licensed tool. To make the learning of JMeter more simple and free, there are many videos created by Raghav Pal which are embedded in the below page:

Many performance testers are more interested in reading rather than watching the videos so they can refer below link where step-by-step topics cover full JMeter learning along with challenges, solutions and some small tips:

Still, you have doubt then you can join the WhatsApp group and clear your doubts:

JMeter WhatsApp Group Link

Dynatrace Official Video for beginners

Dynatrace, an APM tool who provides fabulous support to identify the performance bottleneck, AI-powered analytics, full-stack discovery, pure-path coverage with code-level analysis and many more monitoring and profiling features. Many times I saw in the group chat, people were searching for Dynatrace training which is hard to get (at least nowadays) and then their search is ended with zero.

To help such a people, Dynatrace has published an official video for the beginners who want to start their career in performance engineering field with Dynatrace. This video is also helpful for those who are willing to learn the APM tool and add a new skill in their resume. But yes; practical knowledge is must. So let's start your journey with Dynatrace official video which describes the basics concept and will help to grab knowledge of this wonderful tool.

Thread Dump Analysis - Treadmill Pattern

Thread Dump is a snapshot of the status of all the threads at a particular time. Thread dumps are vital artefacts to diagnose CPU spikes, deadlocks, poor response times, memory problems, unresponsive applications, and other system problems. Thread Dump helps to find out what every thread in the JVM is doing at a particular point in time. A common phenomenon is once a high CPU usage is observed then thread dump analysis is required.

Sometime back, I have shared articles which described Athlete Thread Pattern in which thread is in Runnable state and does not consume CPU. Today, you will see the opposite case of Athlete Thread Pattern that is when the thread is in continuously Runnable state and consumes high CPU. Such type of thread pattern is named as Treadmill Thread Pattern. A simple and descriptive article published in fastThread blog. Happy to share with you to enhance your knowledge repository.ance of AI.

Related Articles:

Why AIOps is required in APM?

Application Performance Monitoring (APM) tools are approaching to next advance level day-by-day. AppDymanics, Dynatrace, New Relic are the big placeholders in this list. Sometime back I have shared an article link in which the author described why application monitoring is as important as application performance testing. Now, today, I am sharing an article link which is published by AppDymanics, described not only application monitoring is enough but it should be enabled with AIOps. What is AIOps and why it is required with the Application Performance Monitoring Tool? Read the article and grab the knowledge:

If you are a performance engineer or on the path to become a successful performance engineer then this article would be helpful for you to understand the import.

2019 Performance Testing Trend - I

I have shared many article links in the past which provide information about the current trends in Performance Testing world. Such articles give an indication of the change in native performance testing process and start a new era where DevOps, AIOps, APM tools will expand their territories and dethroned the existing approach by the sword of more realistic and advanced featured tools and methodologies.

Bob Reselman has published the first part of his article at Neotys blog "2019 Performance Testing Trend" which has some interesting facts about the future of performance testing. Read the article and update your skills as per market demand:

Types of Memory Leakage, Reasons and Resolutions

Memory leakage is one of the common bottlenecks found during performance testing. Ideally, endurance (also known as soak) test is conducted to identify the memory leakage in the application. But sometimes the system fails due to memory issues even in the load test. In such a situation, there could be multiple reasons which make the servers out of memory.

Today, I found a good document in which most common out of memory error types are listed out along with reasons and generic solutions. I would suggest to download and save this PDF because in future when you try to resolve the bottleneck related to memory issue then this document could be a key treasure for you.

Thread Dump Analysis - Atherosclerosis Pattern

Sometime back, I have shared an article on Thread Dump Analysis Athlete Pattern posted at the fastThread blog. That article described how to identify which runnable threads really consume CPU. Following the same topic i.e. Thread Dump Analysis Pattern, today I am going to share another pattern i.e. atherosclerosis. In this article, you will come to know how to identify the BLOCKED thread.

If you are learning performance engineering or involved in it then such type of small tips will help you to enhance your knowledge. The fastThread blog has many informative articles, so read them and grab the knowledge of Performance Engineering.

Article Link

JMeter Runtime Result in Non-GUI mode

JMeter is an open-source tool and nowadays it is in high demand. Day-by-day new plug-ins extend the features of JMeter and make it a stronger competitor for the existing licensed tool. Previously we have seen that the addition of HTML report makes the result analysis detailed and colourful. Now this time Taurus (an open source tool) is enhancing JMeter feature especially when the test is running in Non-GUI mode. What is that feature? Read the full detail in this LinkedIn post shared by Mohd Nazim Siddhique:

LinkedIn Article Link

Type of Performance Test Defect

During Performance Testing, many defects are found which can be termed as bottlenecks. Performance Testing defects are usually different from the functional testing defects because they have moreover occurred at the network or server side. As far as I am confident most of the clients do not have proper defect management for performance testing bottlenecks that lead extra time, effort and money when the bottleneck re-appeared in future releases.

Leave this topic here only because today, I am sharing an article written by Naveen Kumar describing the category of Performance Testing Defects. This article will help you to understand the category and severity of the PT bottlenecks. Refer to the link and get the knowledge:

Thread Dump Analysis - Athlete Pattern

Thread Dump is a snapshot of the status of all the threads at a particular time. Thread dumps are vital artefacts to diagnose CPU spikes, deadlocks, poor response times, memory problems, unresponsive applications, and other system problems. Thread Dump helps to find out what every thread in the JVM is doing at a particular point in time. A common phenomenon is once a high CPU usage is observed then thread dump analysis is required. In the analysis, threads are categorised on the basis of their states. But what happens when a thread is in Runnable state and does not consume any CPU? Have you heard about such type of thread?

In today's funda, we will learn about Athlete Thread Pattern. A simple and descriptive article published in fastThread blog. Happy to share with you to enhance your knowledge repository.

How to reduce Page Load time?

Apart from server performance monitoring and optimization, nowadays clients also want to know how will the application or website perform at the browser level when it goes live i.e. how would be the end user experience on their website. In today's article, the author mentioned the steps to reduce the server response time first byte. Here, he is not talking about the server response time but he is pointing to Page Load Time at browser end. He described some good points to reduce the page loading time in the browser and also given detailed information on 9 different page load speed checking tools which help to identify the client-side issues. These issues may restrict the application to load quickly in the browser and fade-up the end users. The mentioned tools in the article also elaborate on why and which components are the culprit for the slowness of page-loading.

This article could be helpful if you are involved in Performance Engineering work. Because performance improvement is not a task to just improve the performance at the server level, but it is an overall performance optimization which is applied on all three levels i.e. client, server and network.

Server Side Performance Monitoring - Issues, Solutions and Tools

Thread Dump Analysis, Heap Dump Analysis and GC Analysis play an important role during performance bottleneck root cause identification. A sudden spike in CPU, high memory utilization, memory leakage etc. many are causes which need a quick analysis and resolution. Are there any tricks which can help for rapid resolution of these issues?

Ram Lakshmanan has presented a slide show in one of the events in which he explained the server side performance issues and tips to solve those issues by choosing the correct tool. Refer the link to watch the slide show and learn the tips:

Future of Performance Monitoring

In the current world, Performance Monitoring as important as Performance Testing. Many big names like DynaTrace, AppDynamics are continuously working to improve the monitoring features in the APM tool. Even they succeed to make strong tools which are not only monitoring the performance of the system but providing the root cause analysis for the bottleneck (these are very basic features, but they are doing more than these).

AppDynamics team has introduced AIOps concept in the tool and improved the AppDynamics features and add a feather in the cap. A video has been twitted by AppDynamics on 'Future of Performance Monitoring'. I would suggest to stream the video and understand where the performance testing world is going.

Details on AIOps

In today's world, trends are changing so rapidly and if we want to go with trend then we need to adopt the new environment and changes and keep ourself up to date. Sometime back I have shared some articles on AIOps, continue with the same, today I am sharing another informative article on what is AIOps and why it is required to understand (especially in performance monitoring field). AppDynamics team use the AIOps concept and enhance the tool with new AIOps features day-by-day.

Gartner has published this article to spread the knowledge of AIOps. Read it and grab the new era trend.

Apache Kafka, Monitoring and Performance Optimization

I got some requests to brief on Kafka which is highly in demand nowadays. Actually, Kafka is an Apache product which is a freeware (open-source) and used for building real-time data pipelines and streaming app like NetFlix. Kafka is used by some big name like LikedIn, Yahoo, Twitter, Uber etc. You can click here to get the list of companies who use Kafka.

I am thankful to RadhaKrishna Prasad who has written a descriptive article on Kafka from Performance Testing point of view. Refer his LinkedIn article to get the details:

Future of Performance Testing

'Change' is a rule of the world. In 80's, nobody worried about the speed, as the lifestyle was slow. And people bore to wait. In 90's, due to an increase in the number of internet users, the need for performance testing of the website arose which is still continuing in today's world. Now, every year we hear about a new technology or methodology which outdated the existing one. Similarly, in Performance Testing, the conventional method is going to die now. Due to the evolution of AI in performance testing, a quick and accurate result along with root cause of bottleneck can be collected easily.

Tom Smith gathered the thought of 14 performance testing executives about the future of performance testing and tuning which are listed-out in his article. Read the article and predict your future in Performance Testing domain.

Cloud Performance Testing

Cloud computing, a new era's software resources which are shared among the different host. Virtualization and cloud have removed the dependency of local servers. Application, storage and other services are easily deployed on the cloud and can be accessed via Web. Since technology has been changed, so definitely it impacts the other traditional method of development and testing.

RadhaKrishna Prasad had written an informative and to-the-point article on cloud performance testing in which he described cloud, cloud performance testing approach and one case study. So, refer the link to get more details on cloud performance testing.

Open-source Tool Demand

In today's world, trends are changing so rapidly and if we want to go with trend then we need to adopt the new environment and changes and keep ourself up to date. I am continuously watching the interest of companies how much they can pay for performance testing. Yes, there is some degradation in the figures for the paid performance testing tool. In the past and even now some companies or clients do not rely on open-source tools like JMeter. But see the below graph, you will find that some of the open-source tools are searched most on the internet. I am not sure about the current demand, uses and how much clients believe on the paid and open-source tool. But these stats show the start of a new era.

Today, I got another article having a list of Performance Testing open-source tool. Check what are the other performance testing tools apart from JMeter and gather the knowledge at least one among them.

Performance Tuning of Servers

Performance Tuning Guidelines of Various Servers:

Little's Law and Amdahl's Law

As we know that the Little's Law is a base of workload modelling. In performance testing world, Little's Law falls under queuing theory through which the stability of a system can be predicted or measured with the help of processing rate (response time) and arrival rate (TPS). Dynatrace has published an article to provide more basic details on queuing theory using Little's Law and Amdahl's Law. Click the link to read the article.

Subscribe to: Posts (Atom)