Archive for the 'sba' Category
Default configuration may not be best for my demo…
August 3rd, 2007Last night I spoke at the Dusseldorf JUG and had a very nice time with one exception: I broke my last demo. YIPE!
I have suffered for hours since then going over my code and trying to see where I dropped the ball and finally, looking elsewhere, have come upon the reason!
The default setting for replication in the cluster configuration that I selected for the demo in GigaSpaces 6.0 XAP is sync-rec-ack. sync-rec-ack defines the replication behavior to be ‘nearly synchronous’ with the client thread blocking only long enough to ensure that the request has been sent - not received - by the backup space.
I changed the setting to fully synchronous - called ’sync’ in our configuration, and things are much much better. Now, I understand how Virgin Mobile succeeded with our technology where I failed. (they use the sync setting).
For those of you who were watching, my apologies for losing that one order when I killed both of the primary spaces as it was running. With this new improvement, I swear it won’t happen again!
Cheers,
Owen.
JavaPolis 2006 Presentation is online
June 6th, 2007Finally our JavaPolis 06 presentation is online. I promised a link 6 months ago, so here it is: http://www.bejug.org/confluenceBeJUG/display/PARLEYS/SBA+-+Scalable+SOA
Network latency vs. end-to-end latency
May 4th, 2007Geva Perry wrote an excellent blog on Extreme Transactions processing on wall street.
"So basically you now have thousands and thousands of machines buying and selling stocks and other securities from other machines based on extremely complex (and automated) computer models. So it has become a latency game — low-latency, that is."
When it comes to [...]
Why browse when you can simply search?
April 23rd, 2007I had interesting discussions in the past few weeks about the use of Object Graphs and how that approach maps into a space model. It reminds me of the discussions I use to have when I used Versant as an object database and object store in the late 90s. Object Graphs assume a certain hierarchy [...]
“Share Nothing Architecture” redefined
April 6th, 2007"Share Nothing Architecture" is a common pattern for scaling out Web applications.
Generally speaking, the idea is to remove dependency between the scaling units, which in the case of a Web application means the Web Servers. Once these units become independent from each other, you can easily scale them just by adding more units. You can [...]
