Hi Tanmay
Since your questions and my answers would be helpful to the general
Synapse community, I am copying my reply to the user@ list
You are correct. I forgot to bind the reply Destination queue "SampleQ2" with JNID name is Sonic MQ. Now it is working fine.
Good to hear!
We have Sonic ESB, but both development and deployment processes are very
complicated and time consuming. I am looking at some OpenSources like
Servicemix, Synapse and OpenESB. Synapse looks very interesting in terms of
Creating Proxy, load balancing and failover of endpoints. But we need to
evaluate few more Capabilities which are very important for Enterprise wide
deployment. My plan is to use Sonic MQ as the messaging bus and Synapse as the
Service or Mediation bus.
Does Synapse have the following capabilities ?
1) Clustering and High Availability
Yes
2) Governance and Monitoring ( events & alerts ) - any kind of Dash board
The WSO2 ESB built on top of Synapse adds more
capabilities/enhancements, and ships with a graphical admin console and
enhanced JMX support
3) Federation
Other things I like to know are
a) Development and Deployment best practices
WSO2 can provide you with consulting, training etc. as well as
opensource sponsorship opportunities to get any features you require
into the codebase
b) Is there any way to avoid the hard-coding of hostname and port number in the
endpoint element of Proxy configuration
<endpoint>
<address uri="http://pasdevarch1:8080/EJBWebService/EchoBean"/>
</endpoint>
Is it possible to do something like
<endpoint>
<address uri="http://${hostname}:${port}/EJBWebService/EchoBean"/>
</endpoint>
and then pass hostname and port number from the property file in each
environment like DEV, QA, PROD
Not right now, but implementation could be straight forward
c) If we have to deploy 100 Proxy Services in a Single Synapse instance , what
is the best way to configure 100 Prxoy Configurations in single file or one
config file for each Proxy and
then include all 100 files in a main file, start Synapse with the main file.
One small example will be helpful.
Either way would result in the same runtime configuration. There is no
difference in performance etc. Hence this will depend on your
environment, frequency of updates and how they are controlled etc. One
option is to save the configurations into an external
registry/repository and get these from multiple ESB instances on
clusters. Also, some users prefer to check in the configurations into
source control (e.g. SVN, CVS..) for better management, tagging etc to
move from DEV, QA, STAGING and PROD etc..
asankha
