Interview with Sarah Hoffmann
Talks
- Inside the machine room of a world map: PostgreSQL and OpenStreetMap
- Wednesday, 17:20
- Berlin 2+3
Social Media
- Could you briefly introduce yourself?
I’m a software developer, working as a freelancer in software development and consulting for OpenStreetMap (OSM) data. I’ve also been a member of the OpenStreetMap community since 2008. I started out as a contributor but gradually got more involved in development. I am the maintainer of Nominatim (a geocoder), osm2pgsql (a tool to load OSM data into postgresql) and waymarkedtrails (a website showing hiking/biking routes from OSM). I’m also part of the OSMF sysadmin team where I maintain the Nominatim servers.
- How do you engage with the PostgreSQL Community?
Even though much of the development I do involves PostgreSQL, I’ve been just a passive user so far. I’m looking forward to meet the community behind the software.
- What will your talk be about, exactly? Why this topic?
The main topic of the talk is to show how PostgreSQL and PostGIS are used when processing OSM data. We do have some slightly peculiar usage patterns and have developed our own methods of dealing with them. I plan to pick out the most interesting problems we have come across using PostgreSQL and show our solutions.
Being on the maintainer side for OSM software myself, I understand the value of knowing how your software gets used. PostgreSQL has a substantial user base in the OSM community. It’s time to meet.
- What is the audience for your talk?
I hope there will be some PostgreSQL developers in the audience but anybody interested in learning about PostgreSQL in OpenStreetMap is welcome.
- What existing knowledge should the attendee have?
Attendees should have a medium level of experience with PostgreSQL/PostGIS and be familiar with the basics of spatial data processing. It helps to know what OpenStreetMap is but I’ll cover all the necessary technical details needed to understand the talk.
- What is the one feature in PostgreSQL 15 which you like most?
The
MERGE
command looks promising. Although as with every new version, the performance improvements are the most important reason to upgrade.- Which other talk at this year’s conference would you like to see?
Anything related to performance of PostgreSQL. Exploring Linux Memory Usage and IO Performance for PostgreSQL promises to be very relevant for my work with OSM data.
- Which measure, action, feature or activity would—in your eyes—help to accelerate the adoption of PostgreSQL?
The biggest obstacle for OSM data users has always been that they need to get the tuning of the database right. With the wrong parameters a database import of OSM can easily take a week instead of a day. PostgreSQL has come a long way in the last years in terms of being more usable right out of the box. But when things go wrong, it is still extremely hard to find the root cause. Anything that can help novice users understand their setup better would be helpful.