This is something I have seen many times now: a customer calls us because they
lost some data and they want help recovering.
Now you must be wondering: surely if they lost data they can just recover from
their last backup, right? Right — they had that. However we know that
pg_dump takes a while to run and is stressful on the server, so it's normally
run just once a day or so. What happens if you've been running almost a full
work day since your last backup? It's a lot of data to lose.
Come on, admit it: you've always wanted to display the infomask bits from a tuple header in a human-readable manner, but you've never gotten around to it and you still keep htup.h in display while you peek around tuples.
Fortunately, that time is now past! Here's a short and simple recipe to decode the bits for your reading pleasure. Gone is the htup.h cheat sheet. Here's what you need:
We were able to fund two features for PostgreSQL, both of which will hopefully hit for 9.2. The first is work to be done by Greg Smith with pg_stat_statement. The other was fully funded by Heroku which is standardized URI support for libpq and psql. It is my hope that we will continue to use PostgreSQL Conference to actively fund features.
That said, there are changes in the wind. First, PostgreSQL Conference is changing from a semi-annual conference to an annual conference. There is just no way the community can support four north american conference (PgWest, PgEast, PgCon, Postgres Open). What is unknown at this point is whether or not PgWest and PgEast will continue, or if we will just merge them and have PostgreSQL Conference. What is known is that the next conference will not be on the East coast. We are currently negotiating with Denver, Seattle, and San Jose (for a repeat).
So what else is new? Generally speaking, the PostgreSQL conference is operated by a small team within CMD with a few select community members and partners picking up some stray pieces. That has changed this time around. We have an organization and planning committee of 26, all of whom are community members and attendees and/or speakers of the conference. These folks have been an invaluable input to the conference, allowing us to learn from the people that are actually giving us the money to pull this conference off.
Stay tuned for more information in the next couple of weeks on various decisions in regards to the conference!
These are filling up fast, so you will want to get your registration in.
Of note, Jim Mlodgenski maintainer of Stado (a proper, stable fork of GridSQL) will be teaching a Practical PostgreSQL Administration course. This is a full day course. Jim has graciously agreed to allow his percentage of the training revenue to be used for the feature development community initiative.
I received (reprinted with permission) this email from him today:
Hi Josh,
I am excited to say that the Schemaverse contest at DEFCON 19 went great! By the end of the tournament we had 108 registered players and over a million queries ran against the game in a four day span. Not only did the server do great as far as performance goes but the fact that it wasn't exploited during DEFCON is also an impressive stat to note.
I can also proudly tell you that your contribution was mentioned during my own two presentations, found in our How To guide, discussed at our contest booth (right by the front doors to the high traffic contest area! :D) AND was announced during the DEFCON 19 closing ceremonies during my allotted 2 minutes of speaking time.
The winner of your prize is Ian Haken (xxxx@xxxxxx.com). He kicked some butt in the competition and is certainly deserving of it. He has authorized me to send you his name and email. If you need any further details you can talk to him directly.
I likely sound like a broken record at this point but I really do need to say thanks again. Your contribution definitely helped us generate some interest in the first year of our competition and has helped us gain the respect needed to return with the contest for years to come.
Best Regards,
Josh (Abstrct) McDougall
http://schemaverse.com/
It was an honor to sponsor this contest. It is great little things like this that truly show the power of PostgreSQL in places you least expect.
Just a note, although the CFP is technically closed we have not closed the submission form, if you wanted to sneak in a talk or two you are welcome to.
As I have posted in previous articles (
Fixing foreign key deadlocks and
Part 2),
I am working on reducing the lock
strength required by foreign key checks. I have
a working patch
that solves a lot of the problems already; however it doesn't solve the one
problem that I initially set to fix. It turned out to require a bit more
rejiggering than I initially considered.
Note: this article assumes that you know what I have already done in the
patch I posted. If you want to follow through, I suggest you read
the first two links above as an introduction.
May 25th: Talk submission opens August 12th: Talk submission closes (EXTENDED!) August 16th: Speaker notification
Help us continue to provide the overwhelming support to the PostgreSQL Community we always have. Submit your talk today!