Thread: "Serializable Snapshot Isolation in PostgreSQL" presentation at VLDB in Istanbul

"Serializable Snapshot Isolation in PostgreSQL" presentation at VLDB in Istanbul

From
"Kevin Grittner"
Date:
From the abstract that you can also find here:
 
http://sacan.biomed.drexel.edu/vldb2012/program.php#Industrial_Session_Abstracts
 
 
>> Serializable Snapshot Isolation in PostgreSQL
   Dan Ports (MIT CSAIL), Kevin Grittner (Wisconsin Supreme Court)

This paper describes our experience implementing PostgreSQL’s new
serializable isolation level. It is based on the recently-developed
Serializable Snapshot Isolation (SSI) technique. This is the first
implementation of SSI in a production database release as well as the
first in a database that did not previously have a lock-based
serializable isolation level. We reflect on our experience and
describe how we overcame some of the resulting challenges, including
the implementation of a new lock manager, a technique for ensuring
memory usage is bounded, and integration with other PostgreSQL
features. We also introduce an extension to SSI that improves
performance for read-only transactions. We evaluate PostgreSQL’s
serializable isolation level using several benchmarks and show that
it achieves performance only slightly below that of snapshot
isolation, and significantly outperforms the traditional two-phase
locking approach on read-intensive workloads.
 
 
I notice that many of the other papers and demonstrations involve
implementations of the ideas in PostgreSQL.  I'll try to attend those
sessions where possible, and see if any of these ideas seem "ripe"
for core enhancements and free of conflicting IP claims.
 
-Kevin

On 1 July 2012 23:22, Kevin Grittner <Kevin.Grittner@wicourts.gov> wrote:
> From the abstract that you can also find here:
>
> http://sacan.biomed.drexel.edu/vldb2012/program.php#Industrial_Session_Abstracts
>
>
>>> Serializable Snapshot Isolation in PostgreSQL
>    Dan Ports (MIT CSAIL), Kevin Grittner (Wisconsin Supreme Court)
>
> This paper describes our experience implementing PostgreSQL’s new
> serializable isolation level. It is based on the recently-developed
> Serializable Snapshot Isolation (SSI) technique. This is the first
> implementation of SSI in a production database release...

Well done both. This is a major step forwards.

I would very much like it if you described yourself as representing
the PostgreSQL community as well.

--
 Simon Riggs                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services

> I would very much like it if you described yourself as representing
> the PostgreSQL community as well.

+1


--
Josh Berkus
PostgreSQL Experts Inc.
http://pgexperts.com



[resending as I wasn't subscribed to the list when I originally sent
this message; apologies if you get it twice]

You can find the full text of the paper at:
http://drkp.net/drkp/papers/ssi-vldb12.pdf

It doesn't assume familiarity with SSI, so it might be a good place to
start for someone who wanted to learn how SERIALIZABLE mode works (in
great detail). There are also some new benchmark results that may be of
interest.

Dan

--
Dan R. K. Ports              MIT CSAIL                http://drkp.net/

You can find the full text of the paper at:
http://drkp.net/drkp/papers/ssi-vldb12.pdf

It doesn't assume familiarity with SSI, so it might be a good place to
start for someone who wanted to learn how SERIALIZABLE mode works (in
great detail). There are also some new benchmark results that may be of
interest.

Dan

--
Dan R. K. Ports              MIT CSAIL                http://drkp.net/