Personal tools

Document Actions

PostgreSQL Cluster Developers' Meeting

by Nozomi Anzai last modified 2009-09-18 12:53

PostgreSQL Cluster Developers' Meeting
November 19th, 2009 - Tokyo, Japan
Call for participants

Sep. 18th, 2009

It is with great pleasure that The PostgreSQL Conference 2009 Japan Executive Commitee (the Committee) welcome an associated event, the PostgreSQL Cluster Developers' Meeting and is calling for participants as described below. The meeting will be held in English and translators will not be available.

Please contact koich [at] to join. Due to limited meeting space, the Committee might not accept some participants.

1. Situation

A number of PostgreSQL cluster and replication products are available in the market and being improved: Slony, pgpool-II, PGCluster, PgBouncer, PL/Proxy, Postgres-R, PostgresForest, GridSQL, Synch_Rep, Hot Standby and so on. As PostgreSQL is now being deployed in a wide range of business applications, demands for PostgreSQL cluster products seem to be escalating.

Unfortunately, many PostgreSQL users are confused on how do decide which products or projects deliver the functionality and features which meet their needs. This may be partly because there exists no widely-shared guide to PostgreSQL cluster products and no standardized definitions or measures of function and performance for different products. It is also possible that users are confused because no single PostgreSQL cluster product is good enough for all general purposes, covering transparent failover, load balancing, and horizontal scaling.

2. Goal

The first goal of the meeting is to share the view of current/future PostgreSQL cluster products and applications, and hopefully to build a plan for providing a widely-shared guide to these products and applications for educating potential users. A second goal is to share the common requirements for future projects and to see what components, APIs, standards, tools and/or PostgreSQL features would enable or improve several clustering solutions.

3. Participants

Participants should be software developers currently involved in one or more PostgreSQL clustering or replication development efforts, products, or projects.

4. Agenda

The following is a proposal. Any inputs/suggestions are welcome.

1) Current clustering definitions, use cases and customer goals (short)

  1. User goals
  2. Specific use cases
  3. Current market for PostgreSQL clustering
  4. Competitive market of other DBMSes

2) Current PostgreSQL cluster products review

  1. brief (5 minutes) presentation from each developer or team

3) Future requirements and expectations

  1. Common issues to several products
    • Usability
    • Administration
  2. Application or industry specific issues

4) Technical Issuesi in clustering design

  1. Challenges
    • High Availability
    • Scalability (read/write)
  2. Specifications and APIs

5) Plans for future development

  1. a)To be developed in core PostgreSQL
    • Standby/replication (sync/async)/partitioning
    • Transaction Management
    • APIs and interfaces
    • Tools
  2. To be developed separately
  3. Merging clustering projects/products?

5. Schedule

The meeting is scheduled on Nov. 19th, 2009, Thursday. It will start at 9:30AM and close at 5:30PM and be held at Room "B" in the AP Hamamatsu-cho conference hall, where the PostgreSQL Conference Japan will be held next day.

6. Welcome Party

Food and drinks will be served to all participants after the meeting. The party will start at 6:30PM at Mielparque Tokyo, which is near by the AP Hamamatsu-cho conference hall.

7. Sponsors

This meeting is sponsored by NTT Open Source Software Center, cooperated by NTT DATA Corporation, SRA OSS, Inc. Japan.

You are here: Home イベント PostgreSQL Conference 2009 Japan - JPUG 10th Anniversary Conference - English PostgreSQL Cluster Developers' Meeting