sh + database   20

Intel Optane SSD Microbenchmark
When using LMDB, the LMDB engine will never be the bottleneck in your workloads. When you move onto faster storage technologies, LMDB will let you utilize the full potential of that hardware. Inferior technologies like LSM designs won't.
7 days ago by sh
Building a scalable time-series database on PostgreSQL | Hacker News
I really like that Postgres is "good enough" solution for almost every use case by now besides relational data, be it document storage, full text search, messaging - or time series now. Nothing wrong with having less stack to worry about, especially for prototypes and small scale projects!


Loved this article, as a competing database company, they did a fantastic job relating to developers and being authentic! Great job, please keep this up, it will definitely make you a winner.
Key quote of interest: "135K writes per second per node" this is very good! Quite impressed.
january 2018 by sh
How We’re Building a Business to Last | Hacker News
Cloud didn't work because AWS/GCloud have enormous moats of pricing and brand recognition. They drive margins down to epsilon, and if your product sees meaningful adoption in the industry they launch their own service and take all your customers.
misc  database 
january 2017 by sh
Bedrock by Expensify
cross-colo replication on top of sqlite
october 2016 by sh
Immutability, MVCC, and garbage collection at Xaprb
datomic & couchdb started out with "append only" data models. immutability is great, but storage is not infinite.

eventually they needed to do rewrite-the-world compaction of their append-only b-trees. turns out this takes awhile.

mature databases (pg, innodb, oracle, mssql) have been doing this for awhile by using mvcc for *short term* immutability (ACID) but compaction-as-you-go.

becoming a general purpose database likely means datomic & couchdb will have the same underlying data model, regardless of whether it's relational or schemaless. it's all the same bytes.
december 2013 by sh
RocksDB | A persistent key-value store for fast storage environments
key/value store that assumes fast storage (Flash) and databases larger than RAM (in contrast to LevelDB). from Facebook. also used by Flink, can be saved to S3/HDFS easily.
database  programming 
november 2013 by sh
The genius and folly of MongoDB - Bryce Nyeggen's Blog
When you realize “it’s just mmap”, all the BS about optimizing your working set to fit within RAM, the cratering of performance once you hit disk, the fragmentation if you routinely delete or grow records, etc., make perfect sense.
october 2013 by sh
adds aggregate roots to sql, global transactions via TrueTime API
database  google 
september 2012 by sh
DBMS Musings: If all these new DBMS technologies are so scalable, why are Oracle and DB2 still on top of TPC-C? A roadmap to end their dominance.
from abadi, one of the voltdb guys, applies event sourcing, global ordering of transactions, which are atomic/server-side-only. allows a lot of freedom in applying/replicating the events across machines with little coordination.
may 2012 by sh
deterministically pre-order transactions so that replicas can be active/active with no synchronization
database  filetype:pdf  media:document 
september 2010 by sh
WWW SQL Designer
This tool allows you to draw and create database schemas (E-R diagrams) directly in browser, without the need for any external programs.
march 2008 by sh

Copy this bookmark: