Introduction
Hello, I'm Yuto Takashi.
On this blog, I share various insights about introspection (reflection) and engineering.
Today, I'd like to explain SRE in an easy-to-understand way using mobile games as an example.
SRE is Google's proposed approach to how systems should be operated.
Recently, I've been seeing more and more articles about companies launching SRE teams, and I think it's a term that's gradually becoming more common in the IT industry.
What is SRE?
To put it simply, SRE stands for Site Reliability Engineering, a concept proposed by Google.
Specifically, it involves the following:
- Ensuring site reliability
- Taking on dual roles of operations and reliability improvement
- Actively writing code
- Replacing operations with cloud and automation
Let me explain each of these using mobile games as an example.
What Does "Ensuring Site Reliability" Mean?
Site reliability specifically means ensuring that internet-connected services can be used without problems.
In terms of mobile games, this means guaranteeing that the game can be used without issues as much as possible.
Mobile games sometimes become unavailable due to various factors, right?
For example, servers might go down due to traffic spikes during an anime collaboration event, or the game might become unstable after an update.
Every time something like this happens, the people at game companies work hard to get the game back up and running as quickly as possible.
In SRE, there's a responsibility to ensure that users can use the service with peace of mind.
Dual Roles of Operations and Reliability Improvement
Operations refers to the daily tasks performed by system operators.
In mobile games, this corresponds to game maintenance. Mobile games frequently have new events starting and new features being added, right?
During these times, maintenance work is performed internally to enable new events and features.
Additionally, based on past event experiences, measures are taken in advance to prevent server outages, and daily efforts are made to ensure everyone can enjoy the game with peace of mind.
Actively Writing Code
In SRE, rather than just performing operations manually, various tasks are automated to make operations more efficient. Instead of manually recovering when a site goes down, systems are set up to automatically recover when problems occur—improvement through engineering.
For example, in mobile games, when a new version feature is released, code is written to automate delivery to everyone with minimal manual intervention from operators.
Replacing Operations with Cloud and Automation
When the number of mobile game users grows steadily, game company staff increase server counts and enhance server specs to ensure users can enjoy the game comfortably.
If user growth is gradual, operators might be able to handle it by setting up servers manually.
However, when user numbers spike rapidly, performing these tasks manually becomes extremely difficult.
It's possible to design systems that perform such tasks automatically.
This is what SRE professionals do—leveraging cloud and automation technologies to efficiently handle areas that used to require human labor.
Summary
- SRE is the concept of "Site Reliability Engineering" proposed by Google
- Ensuring site reliability
- Taking on dual roles of operations and reliability improvement
- Actively writing code
- Replacing operations with cloud and automation
This concludes the brief explanation of SRE.
For more detailed information, please refer to Google's SRE book.
[📦 商品リンク: moshimo-book-sre-google]
The English version is available for free online. If you're comfortable with English, why not give it a read?
Site Reliability Engineering