Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Why You Shouldn't Write a Database
Search
benbjohnson
October 22, 2015
Technology
2
540
Why You Shouldn't Write a Database
Talk given at the October 2015 Denver Go meetup.
benbjohnson
October 22, 2015
Tweet
Share
More Decks by benbjohnson
See All by benbjohnson
Static Code Analysis Using SSA
benbjohnson
6
1.7k
Writing a High Performance Database in Go
benbjohnson
11
6.7k
Raft: The Understandable Distributed Consensus Protocol
benbjohnson
106
28k
Behavioral Analytics - Understanding the "why" and "how" of your users.
benbjohnson
9
2k
New Tech - Sky
benbjohnson
2
370
Behavioral Databases
benbjohnson
16
2.5k
Sky Analytics Pitch Deck
benbjohnson
0
580
Other Decks in Technology
See All in Technology
OPENLOGI Company Profile for engineer
hr01
1
20k
(機械学習システムでも) SLO から始める信頼性構築 - ゆる SRE#9 2025/02/21
daigo0927
0
240
Amazon Aurora のバージョンアップ手法について
smt7174
1
110
わたしがEMとして入社した「最初の100日」の過ごし方 / EMConfJp2025
daiksy
13
3.6k
脳波を用いた嗜好マッチングシステム
hokkey621
0
260
コンテナサプライチェーンセキュリティ
kyohmizu
1
130
偏光画像処理ライブラリを作った話
elerac
1
150
ExaDB-XSで利用されているExadata Exascaleについて
oracle4engineer
PRO
3
150
大規模アジャイルフレームワークから学ぶエンジニアマネジメントの本質
staka121
PRO
2
150
The Future of SEO: The Impact of AI on Search
badams
0
260
あれは良かった、あれは苦労したB2B2C型SaaSの新規開発におけるCloud Spanner
hirohito1108
2
880
OSS構成管理ツールCMDBuildを使ったAWSリソース管理の自動化
satorufunai
0
390
Featured
See All Featured
Into the Great Unknown - MozCon
thekraken
35
1.6k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
jcasabona
32
2.1k
The Pragmatic Product Professional
lauravandoore
32
6.4k
Creating an realtime collaboration tool: Agile Flush - .NET Oxford
marcduiker
27
1.9k
Keith and Marios Guide to Fast Websites
keithpitt
411
22k
Done Done
chrislema
182
16k
JavaScript: Past, Present, and Future - NDC Porto 2020
reverentgeek
47
5.2k
GitHub's CSS Performance
jonrohan
1030
460k
[RailsConf 2023] Rails as a piece of cake
palkan
53
5.3k
Optimizing for Happiness
mojombo
376
70k
Unsuck your backbone
ammeep
669
57k
Design and Strategy: How to Deal with People Who Don’t "Get" Design
morganepeng
129
19k
Transcript
Why You Shouldn’t Write a Database Ben Johnson
Let’s define “database”
low-level LevelDB, BerkeleyDB, LMDB
Direct interface to OS Files, pages, & blocks Responsible for
data integrity Typically key/value
high-level SQLite, MySQL, PostgreSQL, Mongo, InfluxDB, etc
Builds on low-level stores Rows, tables, indexes Interfaces with end
user Relational, document, time series
Me
SkyDB ReportifyDB BoltDB InfluxDB
There are n+1 reasons not to write a database
#1. High barrier to entry
How to write a database
None
legit
sorta, but not quite
seriously?
suggests using XML
nope
Available resources?
Research papers!
Research papers! Narrowly focused (indexing, storage, locks)
Research papers! High level (little or no code)
Research papers! Assumes a Ph.D
Read source code!
Read source code! MySQL >1M SLOC
Read source code! Even small databases are 10KLOC+
There is no Writing Databases 101
#2. Debugging Sucks
Bugs are catastrophic Cause corruption, loss of data integrity
Bugs are catastrophic Users with data loss are very unhappy
Find a good hex editor
Find a good hex editor (I use Hex Fiend)
None
This is not what a database looks like
None
THIS is what a database looks like
Debugging w/o data Users usually can’t release their data
#3. Tradeoffs
Users don’t understand tradeoffs
None
Read Optimized vs Write Optimized
CPU Bound vs IO Bound
https://gist.github.com/jboner/2841832
Features are liabilities It’s not if you have bugs, it’s
how many
#4. Limited Community
Very few people who have written a production database
TONS of people who will tell you why your database
sucks
Databases are hard
The End
The End (Just kidding!)
Not scared off yet?
Only 2 reasons to write a database:
#1. To learn
Data Integrity Indexing Isolation levels Recovery Prefetching Parser & Lexers
Replication Snapshotting Transactions Materialized View Referential Integrity Query Compilation Query Planning Query Optimization Serializability Write Ahead Log Memory Profiling Redo/Undo Log Snapshot Isolation Recovery Two-Phase Commit Quorums
#2. Gain efficiency (For a specific use case)
The more generic your database, the fewer assumptions you can
make
Go write a database!
The End @benbjohnson
Questions? @benbjohnson