You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@qpid.apache.org by "Ted Ross (JIRA)" <ji...@apache.org> on 2014/01/30 19:20:17 UTC
[jira] [Closed] (QPID-5063) Execute a program when a message
arrives addressed to a specific adress
[ https://issues.apache.org/jira/browse/QPID-5063?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Ted Ross closed QPID-5063.
--------------------------
Resolution: Duplicate
> Execute a program when a message arrives addressed to a specific adress
> -----------------------------------------------------------------------
>
> Key: QPID-5063
> URL: https://issues.apache.org/jira/browse/QPID-5063
> Project: Qpid
> Issue Type: Improvement
> Components: Qpid Dispatch
> Environment: linux
> Reporter: Sean Gallagher
> Assignee: Ted Ross
> Priority: Minor
>
> Start a server (if not already running) when a message arrives at a pre-configured address.
> Broadly similar in intent to xinetd or http CGI scripts.
> My application has some large jobs that execute infrequently. It would be good if these could be run on demand rather than keeping a large process in memory that does nothing most of the time. Having the server restart when needed is also a good guard against resource leaks / selfish garbage collectors.
> There doesn't seem to be a convenient way to do this at present but I'm thinking it might be added relatively easily to Dispatch Router and I suspect many would find it useful.
> Working out locks / race conditions would be tricky but one simple idea I had would be to execute a user-supplied command line when a message arrives and then lockout further invocations for a configurable time period. This would give the new server time to startup and connect to Dispatch. Dispatch would not invoke the command if a link was available to deliver the message.
> A section could be added to qpid-dispatch.conf to configure this. e.g.
> on-demand {
> ##
> ## deliveries to this address will trigger the command if link does not already exist
> ##
> to: myserver/myservice
> ##
> ## command line to run
> ##
> run: myprogram
> ##
> ## delay after triggering the command before it may be retried (in response to a new delivery)
> ##
> lockout-interval: 60
> }
> This idea could be further developed in many directions, what I've put here is the basics.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@qpid.apache.org
For additional commands, e-mail: dev-help@qpid.apache.org