How to deal with excessive latency and performance issues.
In the drive to deliver highly scalable services that meet the demands of mobile users, enterprises and, increasingly, the Internet of things, the software underpinning them has become incredibly complex. At the same time, the systems that run these complex workloads have become prone to troublesome and often baffling performance issues.
Response times measured over the course of millions of user requests can vary dramatically. Most requests will complete within the time needed to sustain acceptable quality of service. But that statement conceals a problem. Plots of the distribution of response times often show a long tail. For the “tail” requests, the latency experienced by the user is excessive.
To read more, click here.
 |
 |
 |
 |
 |
 |
 |
 |
 |  |
Leave a Reply