Skip to main content

Drop in application throughput on AWS (Mystery Unresolved)

I have posted a question on stackoverflow about sudden reduction in application throughout. I also follows up on this conversation with AWS support and Following is the gist of conversation with AWS Support -
tl dr:
AWS instances my team was using was not set with enhanced networking capabilities to be able to get max n/w performance (i.e. 10 gigabits on c4.8xlarge instance). For example ixgbevf on test-aws-am was not set to 2.14.2.
Long Version:
The test in question is a static html page. GET request, no complicated logic, No EBS, DJ-CTS etc.
  • What n/w capabilities should we experience when instance does not satisfy enhanced networking capabilities?
    AWS Support: We really do not have specific numbers because that varies... by a lot of factors (time of day, other instances sharing network in same location.. and several other factors... What we do, is to advice customers to do benchmarking tests to confirm that the instances meet the performance expectations... for the applications
  • During the test 18430 KiloBytes/sec data is transferred which is way under the limit of 10 gigabits n/w.
    AWS support has been insistent on probability of throughput throttle being an application issue.
    AWS Support: Through our testing we have eliminated ELB as a potential bottleneck for the drop in throughput and we know that the issue is occurring at the back-end instances. Looking at the back-end instances, the general performance metrics such as CPU utilization, Network In, Network Out etc for both the instances looks good. This indicates that the issue we are facing could be an application issue. could you please add another instance and try the test again?
    I can not confirm on this since there have been zero errors during multiple iterations of test runs. Once tests hit the lower limit on throughput then any subsequent test run shows the results in the lower range of throughout (about 8000 requests/sec) but if I wait for few hours (about 2 to 3 hrs) and run test then it is back to same behavior that is higher throughput for about half hour and then back to 8000 requests/sec. I excluded the possibilities of adding another instances as testing without ELB (described below) exhibits same behavior.
  • ELB Prewarm did not help on addressing reduced throughput. Moreover warm up is time bound and it scales down after some time. The warmup rather showed the more skewed results and throughput dropped in 5 minutes than without it - 



Before Warmup >>


After Warmup >>



  • Different instance type for test agent from m4.4xlarge to m4.10xlarge and c4.large have shown same behaviour of throttle in application throughput. Test runs with ramp up period of increasing threads hold same behavior barring that it reaches the higher limit after greater amount of time attributed to the ramp up period and not half an hour
  • Testing without ELB exhibits same behavior except that the high throughput would be in the range of 8000 requests/sec and then it drops to about 4000 requests/sec. Despite tests use custom DNS resolver; this further excludes any anomaly caused by probable DNS caching of ELBs.

Mystery of drop in application throughput was never solved.

Later on I got to know from in house n/w engineers that it is not only enhanced n/w capabilities but also placement group which is required to be able to use 10 gigabit n/w performances. Unfortunately placement group is limited to having instances in same availability zone. Is not it risky?

Popular posts from this blog

Distributed Load Testing with JMeter

Distributed Testing with JMeter When one JMeter client is not able to offer amount of threads required for load testing then distributed testing is used. In distributed testing - One instance of JMeter client can control number of JMeter instances and collect data from them Test plan does not need to be copied to each server, the client sends it to all servers note - JMeter will run all the threads on all the servers, hence 100 threads on 5 JMeter server would pump 500 threads in total. If many server instances are used, the client JMeter can become overloaded and so the client network connection. This has been improved in latest versions of JMeter by switching to Stripped modes, but you should always check that your client is not overloaded When Client (master) and Server (slave) nodes are on same network (no SSH required) Configure Client Node Herein client is referred as the machine controlling test execution on other JMeter nodes. This is also referred

Appium and android mobile app automation

Next appium and Android mobile app automation video tutoria l is live. If you are new to appium then please check - appium-tutorial This video tutorial covers - Start vysor (Just for this session and not mobile automation :)) Start appium and start appium inspector Desired Capabilities platformName - Android deviceName - L2N0219828001013 (as seen on "adb devices") Saved Capability Sets Start Session Scan app elements using appium inspector Get appPackage and appActivity using "APK info" app Install "APK info" app and open app whose appPackage and appActivity are required i.e. calculator Check top section of app icon com.android.calculator2 is app package com.android.calculator2.Calculator is app activity testng.xml file settings for running Android app tests Test details com.seleniumtests.tests.mobile.AndroidAppTest and CalculatorScreen class View beautiful STF test report  

Verify email confirmation using Selenium

Note: If you are new to java and selenium then start with selenium java training videos .     Email confirmation seems to be integral part of any registration process. I came across an application which lets you provide your email address. You can follow the sign up link in you mail and then complete the registration process. Lets consider we provide GMail address for it. Now if were to use only Selenium then we would have to follow following steps - Launch GMail using Selenium; Some how search for new mail in the list of available mails; Some how click on it; Parse the mail message; Get the registration link; Follow up with registration process What do you think of an approach in which you can