Wednesday, March 18, 2009

QCon Presentation Slides For Download

I presented Next-Generation AMQP Messaging Performance, Architectures, and Ecosystems with Red Hat Enterprise MRG last week at QCon London, along with Carl Trieloff.  The talk included a detailed use case of work we've been doing with customers to build financial trading systems and stock exchanges with MRG.  Messaging deployments don't get any higher end than this, so it's an interesting study. 

In particular, we've added capabilities in MRG like LVQ, Ring Queue, and TTL so that companies can build market caches or provide updated streaming quotes directly from the messaging broker rather than build these capabilities themselves.  We've also done a lot of work to get some pretty impressive performance (e.g. fully reliable latency across three peers in around 60 microseconds).  And, of course, we support capabilities like active/active clustering for high availability and federation for disaster recovery.

Beyond the case study, we also talked about the new ecosystems that are forming around AMQP and MRG as well as new types of messaging-oriented applications people are building now.

You can download our slides from the talk here.

5 comments:

Trustin Lee said...

Thanks for sharing the slides. I wonder how's the performance of the Java version comparing to the C++ version? The slides seem to demonstrate C++ version only.

Anonymous said...

It seems to me that the AMQP Broker becomes a bottleneck in this solution since every message must be sent and received through the Broker. Has any thought been given to other solutions (e.g. OMG's Data Distribution Service) in which destination information is exchanged, but the actual messages are sent directly to the destinations via either unicast UDP or multicast UDP? Could you comment?

Bryan Che said...

A brokered architecture enables us to offer many higher-level capabilities that are quite valuable. For example, many of the features in the trading use case in the slides (LVQ, TTL, etc) wouldn't be possible without a broker.

The trick is to make sure we can deliver these features with extreme performance. The performance and scalability of the broker we've benchmarked to date are pretty impressive. For example, see our whitepaper from a benchmark we performed in Intel's low-latency lab last year (https://inquiries.redhat.com/go/redhat/MRGWhitepaper?sc_cid=70160000000HhJiAAK). On an 8-core box with 1GB ethernet, we scaled to over 6m OPRA messages/second fanning out to 60 clients.

We've got some updated data we'll be putting out soon that shows we've increased quite a bit since then. And, of course, there's the latency data included in the QCon slides and my other post about RDMA.

Unknown said...

It seems to me the AMQP Specialist is a bottleneck within Cheap Diablo 3 Itemsthis solution because just about every information has to be directed in addition to been given with the Brokerage. Has any considered gained for some other solutions (at the.gary. OMG's Data Circulation Services) during which vacation spot facts are sold, even so the true mail messages are usually directed directly toGW2 Gold the spots by means of both unicast UDP or even multicast UDP? Might you opinion?

summerlijia said...

WoW gold for sale! Guys, do you need any WoW products. Click the colored words to getcheapest WoW gold with fast delivery speed. Hope you a nice day.