Framework For Evaluating Web Data Streaming Servers

I think we need something similar to the CEP framework for Web Data Streaming Servers. It’s particularly hard from the web sites and marketing speak to identify which product is the leader of the pack – although I do have a few :)

Some of these web data streaming product companies are just hurting themselves by not offering sensible demo’s from their own web sites. Or they have a demo, but it doesn’t work outside trading hours :( Others have painful installs and samples which just infuriate developers. Then there is the marketing documentation that says the product support certain technologies, but the download samples don’t have appropriate samples.

Support is an important factor in the uptake of a product. As an example, one streaming product company has offered to install their Blackberry demo on my Blackberry. The same company is also building me a Silverlight sample – they should have shipped it in the product download, but at least support is rectifying the situation.

Marketing is often a problem with product companies – the wrong message can impact the product uptake. I can think of at least one streaming server company that has been impacted by this.

Here are a coupled of areas I look at when evaluating a streaming server:

  • Clustering and failover
  • Consistency of API
  • Easy of subscription/message
  • Performance – throttling, etc
  • Quality of documentation
  • Support and company personal
Advertisement

~ by mdavey on September 14, 2009.

3 Responses to “Framework For Evaluating Web Data Streaming Servers”

  1. How about http://www.lightstreamer.com/?

    • Sorry, not sure what you mean. Do you mean what do I think of Lightstreamer?

  2. For CEP, besides the paper you mentioned, I think it is worth mentioning also STAC Research (http://www.stacresearch.com/).

    For the Web Data Streaming Servers it is a real need to define a benchmarking methodology to distinguish between the enterprise push solutions and many ad-hoc push implementations that don’t stop to appear. I think performances are the most important part of a Streaming Server: creating a push server is a question of weeks, but creating a high performance scalable streaming server is a very sophisticated task.

    At Migratory, we spent many months only on benchmarking and performance tuning. Here is a document explaining the benchmarking methodology and the results obtained with Migratory PushEngine;

    http://www.migratory.ro/data/Migratory%20PushEngine%20-%20Benchmarks.pdf

    Migratory will be happy to make available its platform Migratory PushEngine to an independent company such as STAC Reasearch interested to specialize in Push Servers benchmarking. Migratory will be also interested to contribute at the definition of a benchmark specification for push servers and invite by this channel any expert/company that produces enterprise level push servers to contact us to work together on defining such a specification.


    mihai.rotaru {AT} migratory.ro
    http://www.migratory.ro

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Connecting to %s

 
Follow

Get every new post delivered to your Inbox.

Join 272 other followers