17
Accelerate: State of DevOps 2019 | How Do We Compare?
5 Availability is not included in our cluster analysis because availability measures do not apply the same way for soware solutions that are not provided
in the form of services, such as packaged soware or firmware.
6 Teams can define their availability goals using Service Level Agreements (SLAs) and Service Level Objectives (SLOs) and measure their performance using
Service Level Indicators (SLIs). For more information on developing SLAs, SLOs, and SLIs, you can check out Site Reliability Engineering: How Google
Runs Production Systems (2016) by Beyer et al.
Many professionals approach these metrics as
representing a set of trade-os, believing that
increasing throughput will negatively impact
the reliability of the soware delivery process
and the availability of services. For six years in
a row, however, our research has consistently
shown that speed and stability are outcomes
that enable each other. Cluster analysis of the
four soware delivery measures in the 2019
data reveals four distinct performance profiles,
with statistically significant dierences in
throughput and stability measures among
them.
5
As in previous years, our highest
performers do significantly better on all four
measures, and low performers do significantly
worse in all areas.
In addition to speed and stability, availability
is important for operational performance.
At a high level, availability represents an
ability for technology teams and organizations
to keep promises and assertions about
the soware they are operating. Notably,
availability is about ensuring a product or
service is available to and can be accessed
by your end users.
6
Availability reflects how well teams define
their availability targets, track their current
availability, and learn from any outages,
making sure their feedback loops are
complete. The items used to measure
availability form a valid and reliable
measurement construct.
我的内容管理
展开
