From PostgreSQL wiki

Revision as of 16:01, 17 June 2014 by Markwkm (Talk | contribs)

Jump to: navigation, search


PDXPUG Day 2014

PDXPUG is the Portland PostgreSQL Users' Group.

Date: Saturday, September 6, 2014

Time: 9:30am through 5:30pm

Place: Portland State University, Room EB 102, 1930 SW 4th Avenue, Portland, Oregon 97201. The venue is reachable by street car, bus or light rail, see Tri-Met for schedule and fare information.

Registration: Space is limited, please RSVP! There is no registration fee, but donations to PGUS are appreciated.

Food: We'll provide light snacks & beverages. Breakfast and lunch is on your own. There's a wonderful cart pod right across the street; bring $10-15 for lunch.

Presentations: We are looking for presenters for the PDXPUG Day. Talks should be about 45 minutes in length, and about any of the following topics:

  • PostgreSQL administration and performance
  • Case studies of interesting uses of PostgreSQL and PostGIS
  • Interesting applications built on PostGIS or PostgreSQL
  • Database and/or geographic application development
  • Database-related DevOps
  • SQL and stored procedure programming
  • New Postgres/PostGIS features and hacking Postgres/PostGIS

We will also be taking lightning talks of 5 minutes each, on similar topics. Email your talk proposal to Mark Wong at markwkm -at- postgresql.org.

Special Thanks

We couldn't have done it without help from Portland State University Maseeh College of Engineering and Computer Science, and Datalab, the PSU Data and Information Management Laboratory:

Psu mcecs 4cp.jpg



(subject to change)

6 hour long presentations will be scheduled.

9:30am Introductions Joshua Drake
10:00am Josh Berkus
11:00am Eric Hanson
12:00pm Lunch
1:30pm “Data Near Here” (Reprise): Implementing A Data Search Engine in PostgreSQL Veronika Megler
3:30pm HSTORE, XML, JSON, JSONB OH MY! David Wheeler
4:30pm Lightning Talks

Talk Descriptions

Lightning Talks

Anyone can give a 5 minute presentation! You can decide on the day of the presentation. Just let Mark (markwkm -at- postgresql.org) or Gabrielle know and we will be sure to fit you in.

  • PDXPUG by Chelnik - Information on the Portland PostgreSQL Users Group

“Data Near Here” (Reprise): Implementing A Data Search Engine in PostgreSQL (Veronika Megler)

We’ve heard a lot about Big Data. Scientists are creating data archives that are terabytes, petabytes and even brontobytes in size. But now they’ve collected all this data, how do they find what they’re looking for? Current techniques quickly fail.

“Data Near Here” (DNH) is a search engine, built on PostgreSQL, that uses Internet search-like techniques to solve (parts of) this problem. This presentation describes the app, which is now in production over an observational archive; it also describes some of the techniques used and challenges encountered during its development.

Veronika recently completed her PhD in Computer Science at Portland State University. The research supporting DNH is described in her dissertation. Her core expertise is in adoption of emerging technologies. Her career has taken her into almost every corner of the computer industry, including operations, application development, systems programming, systems management disciplines, project management, and IT management consulting. Fun facts: She’s famous for something she did as an undergrad – writing a cult computer game, in Assembler. Her first computer was an electronics magazine project c.1980 that she built with a soldering iron, using individually-bought capacitors and resistors. (Really).


There has been a lot of work on the representation of unstructured data in PostgreSQL, culminating in the addition of the JSONB type in the forthcoming 9.4 release. JSONB complements the existing HSTORE, XML, and JSON types, not to mention arrays. With so many options, which do you use? As usual it depends on your use case.

In this presentation, we'll review the unstructured data types in PostgreSQL, and look at their advantages and disadvantages for:

  • Document storage
  • Configuration management
  • A “schemaless database”
  • Object serialization
  • Entity/Attribute/Value models
  • Path queries
Personal tools