Hacker News Reader

billoneilon Show HN: I made a MailChimp alternative that connects to your database

Hosted Looker and many other tools are getting direct DB access these days and I'm sure you use products that do it. As an engineer I'm not a fan of it and it can often cause problems so it's best to point them at read replicas but that is not always how companies set them up.

billoneilon Show HN: After 2.5 years on my side project, it has hit £500/month revenue

Most employment contracts have a section where you declare your existing intellectual property. You can also have this amended after you start working there.

billoneilon Using Elasticsearch as the Primary Data Store in an ETL Pipeline

Either way, something still seems off. If MongoDB and Elasticsearch were 3-5x more performant I would still find that surprising. A > 10-20x difference really seems like a configuration or implementation issue.

billoneilon Using Elasticsearch as the Primary Data Store in an ETL Pipeline

> The reason PostgreSQL results were omitted is no matter what kind of optimization we throw at it, the benchmark always took more than 2 hours, regardless of partitioning, whereas MongoDB and Elasticsearch took a couple of minutes.

This seems like something was done incorrectly the comparison shouldn't be that drastic.

> Just one PostgreSQL 9.6.10 instance (shared_buffers = 128MB)

This looks way too low. The postgresql docs say a good starting point for shared_buffers is 25% of the server's memory. In this case that would be 32GB.

https://www.postgresql.org/docs/9.1/runtime-config-resource....