PgConf.Russia 2019

Brendan Gregg’s USE (Utilization, Saturation, Errors) method for monitoring is quite known. There’s also Tom Wilkie’s RED (Rate, Errors, Durations) method, which is suggested to be better suited to monitor services than USE. I want to talk about how we employ these methodologies when we develop our Postgres monitoring in okmeter.io.

At pgconf’17 I talked about our analytics systems based on PostgreSQL. Afterwards we looked at hadoop, s3, presto, vertica, and other frights. Finally we stopped to suffer nonsense and just completed PostgreSQL with ready Greenplum and Clickhouse. As a result, we achieved amazing performance, fast migration, easy maintenance, reliability and horizontal scalability. We enabled to recover the system after fault in two commands, decreased infrastructure costs and expanded functionality due to ANSI SQL, MPP and In-memory. All within the open-source and full SQL paradigm. We called the product GreenHouseSQL, which is our inner whole cycle data platform. In the talk we will show the beauty of solution internals, explain the advantages and flaws, tips and tricks of starting with Greenplum, as well as why do we need Clickhouse, what is left to PostgreSQL, and eventually how does it all work.

While doing development of one of our project we were asked to build HA database using Postgres, geographically distributed.

First our choice was obvious, we started to work with big 3 cloud providers, but soon it was quite understand that everything costs big enough for us. Also there were a bunch of incompatibilities with unsupported extensions as well as londiste replication we were heavily used.

I will talk about why we chose patroni, what types of problem we faced with and patroni's special features can dramatically simplify deploy and everyday usage.

This talk will contain 3 parts: 1. Express PITR setup to the Cloud 2. Latest advancements in WAL-G for backups 3. Why you may need or should avoid this new features, depending on your specifications and workload.