New-age Transactional Systems - Not Your Grandpa's OLTP
John Hugg discusses high volume transaction processing applications with high and low frequency profiles, and how VoltDB can be used for that purpose.
The content has been bookmarked!
There was an error bookmarking this content! Please retry.
Posted by Sebastien Auvray on Aug 20, 2008
In the MRI world, we use memcached, which is a Ruby interface to the libmemcached C client. It is one of the only Ruby libraries that currently supports consistent hashing, which is a requirement for our production environment. When preparing for JRuby, we had a different set of options. One was writing our own hooks into libmemcached via JNA, JRuby FFI, etc. The other was to look for another cache solution in Java, and Ehcache was definitely on top of the list.Would you consider using EHcache with JRuby?
Case Study: IBM's Agile Transformation
Using Drools? See what you're missing! Get the Power of Drools with the Assurance of Red Hat
Monitor your Production Java App - includes JMX! Low Overhead - Free download
agility@scale eKit: 10 Principles, Scaling Model, Metrics, Collaboration
Another option for using consistent hashing with memcached and JRuby would be to use the spymemcached client. The spymemcached client is an asychronous single-threaded client that has proven to work extremely well in the hibernate-memcached project. I have also used the spymemcached client in other projects and its performance was outstanding.
Hi,
As you know, the 'page caching' is by far the fastest caching system in Rails (like 2 orders of magnitude faster than hitting a fragment cache). But unfortunately, standard Rails page caching tends to be harder on J2EE serveurs that sometimes behave really differently from an Apache front end for instance. Especially there is an issue with static files colliding with dynamic requests pathes (just normal for standard rails page caching) but that are redirected agressively on J2EE servers unless you hack the app a little bit specifically for a J2EE deployment.
Also, file based page caching is hardly convincing, even if J2EE servers have a memory cache of the static files. Indeed, there can be latency, disk access, it's not very easy to scale in cluster or to have something close to transactional.
EH Cache would be a very great fit for all that. And Ideally it would be as transparent as possible so we avoid to much J2EE specifics here. Still, as serving pages directly from J2EE server tends to be two orders of magnitude faster than entering a JRuby+Rails runtime, it would be really great to tweak the JRuby-Rack servlet filter so that EH-Cached cached pages get served from directly from that servlet filter without even hitting the JRuby on Rails runtime! It would be a bit like the current file base page caching system but I think this would be faster, easier to scale and won't be screwed up with path collision. Any thought?
Raphaël Valyi.
Thanks for the link Ray !
Haven't had a chance to work with Spymemcached, but have heard of it. It's definitely on my list, and am going to check it out soon.
Hey Raphael ! That's a great idea, and would love to get a memory option for page caches into JRuby/Rack at some point. I'm following the thread you and Nick Sieger are having in jira.codehaus (2832), and will chime in when I get a chance :) Thanks !
Indeed it would definitely be an interesting option!
John Hugg discusses high volume transaction processing applications with high and low frequency profiles, and how VoltDB can be used for that purpose.
Kevlin Henney examines code samples to see what can be learned from them starting from the premise that one won’t write great code unless he knows how to read it.
Jason Ayers share the observations he made watching a team of developers collaborating in real time on the same code base, pushing XP, pair programming and continuous integration to their extremes.
Michael Snoyman presents Yesod, a web framework written in Haskell and containing a web server, templating, ORM, libraries (templating, gravatar, etc.).
Richard Kreuter and Kyle Banker on how to avoid classical RDBMS transactional systems by using compensation mechanisms, transactional messaging or transactional procedures.
Attila Szegedi talks about performance tuning Java and Scala programs at Twitter: how to approach GC problems, the importance of asynchronous I/O, when to use MySQL/Cassandra/Redis, and much more.
One category of risk that project teams need to ensure they address is business value failure – delivering a product that fails to provide value for the business investor.
InfoQ spoke to the authors of Software Systems Architecture on a couple of new topics, the System Context viewpoint and Agile, which have been added to the second edition.
5 comments
Watch Thread Reply