Response Time SLA

Percentiles, histograms, and missed SLA counters

Response time (or latency) percentiles are a ubiquitous measure of system performance. The P999 (99.9th percentile) is a high bar, so it’s a good metric to determine an SLA. For example, my team runs a system with an 800 millisecond response time SLA determined by the P999. To meet that SLA, 99.9% of all requests must take <= 800ms. The system meets the SLA, but here’s its P999 graph: Read more...

Jun 29, 2019

#metrics #percentiles #histograms #sla

2019-04-12 Path and Dimension Metrics

2019-03-25 3 Golang Error Types

2018-12-16 3 Golang Channel Red Flags

2018-12-02 Component-based Software Design

2018-07-22 When MySQL Goes Away

2018-05-13 Thinking in Go

2018-02-04 Bootloader Design Pattern

2018-01-28 Database Scalability: Contention and Crosstalk

2018-01-23 Crash-safe MySQL Replication

2017-12-31 Tech Workers Are Good People

2017-12-10 Approaching a New Software Project

2017-11-12 Ideas, Leaders, and Engineers

2017-09-17 Five-Point Checklist for Excellent Software

2017-09-05 Designing Tools for Integration

2017-08-20 How To Test the Database

2017-08-01 Design Before Implementation

2017-06-17 Reading os/exec.Cmd Output Without Race Conditions

2017-06-11 Custom MongoDB Replica Set Write Concern