:) Re: QPID-5577... Sort of... it's all really because the ConfiguredObject stuff was only ever half completed... the idea always was that the implementation objects would be the thing that got managed... also as I work on AMQP 1.0 management it points out a few other things that need to get tidied up and sorted (like allowing proper hierarchies of types rather than pretending all queues are the same, for instance).
There's also some work that Keith and Alex are doing that may impact - with the basic model around vritualhosts likely to undergo a little bit of change as they build a more generic abstraction for HA clustered vhosts. What I would like to do is to enable everything through AMQP 1.0 management as soon as possible (even over 0-9 and 0-10 protocols)... and if we can build a nice web front end using your javascript implementation of Proton and WebSockets... well, that would be lovely :-) -- Rob On 23 February 2014 17:06, Fraser Adams <[email protected]>wrote: > Can you hear the sound of my heart sinking :-p > > Is this related to JIRA QPID-5577 > > Frase > > > On 23/02/14 15:53, Rob Godfrey wrote: > >> Great... >> >> as a bit of a heads up, I'm likely to be changing stuff a bit more in this >> general area soon... I'm trying to tidy things up so they are a bit more >> consistent and allowing a more generic system for having multiple "types" >> of a given category of configured object... I'll do my best to ensure >> things compile and warn you if I have to change anything. >> >> Cheers, >> Rob >> >> >> On 23 February 2014 16:39, Fraser Adams <[email protected] >> >wrote: >> >> Thanks for this Rob, finally got round to trying this. Although it >>> compiled OK after your updates it was still a little broken, it turned >>> out >>> to be to do with the ExclusivityPolicy changes where the code compiled >>> but >>> i ended up getting ClassCastExceptions 'cause I was casting to Boolean >>> based on the previous assumptions. >>> >>> I've updated with code that explicitly checks ExclusivityPolicy.NONE and >>> it's tickety boo now. >>> >>> I've also had to tweak code in the Broker Object - unfortunately the QMF >>> "port" property for Broker says "TCP Port for AMQP Service" and having >>> added the WebSocket Transport I was returning that instead. I need to >>> think >>> of some way to get some of the more detailed info from the JavaBroker >>> without breaking QMF too, such fun :-) >>> >>> Thanks for sorting out the other stuff - it all seems to be working again >>> now. >>> >>> Frase >>> >>> >>> On 18/02/14 21:08, Rob Godfrey wrote: >>> >>> OK - it now compiles :-) I'm not sure how I'd go about checking that it >>>> actually works but the only thing that there is even a tiny chance of >>>> breakage with is the queue creation... >>>> >>>> https://svn.apache.org/r1569525 >>>> >>>> -- Rob >>>> >>>> >>>> On 18 February 2014 22:02, Fraser Adams <[email protected] >>>> >>>>> wrote: >>>>> >>>> Thanks Rob, >>>> >>>>> I took a quick look and came to the same conclusion. If you're able to >>>>> have a go at it that'd be great, it'll save me from the wrath of my >>>>> significant other :-) >>>>> >>>>> Cheers, >>>>> Frase >>>>> >>>>> >>>>> On 18/02/14 20:27, Rob Godfrey wrote: >>>>> >>>>> Hi Fraser, >>>>> >>>>>> that would have been me... sorry... it was after the 0.26 branch... >>>>>> >>>>>> It'll be fairly easy to get in line for the 0.28 release, the only one >>>>>> that >>>>>> isn't entirely trivial is the createQueue one ... the other two issues >>>>>> are >>>>>> that the Subscription interface has been renamed Consumer and that >>>>>> rather >>>>>> than lifetimePolicy == LifetimePolicy.AUTO_DELETE that should probably >>>>>> now >>>>>> read lifetimePolicy != LifetimePolicy.PERMANENT. Even the createQueue >>>>>> on >>>>>> is just changing the method to have the appropriate values in a map... >>>>>> >>>>>> If I get a chance I'll get it to the point where it compiles tonight. >>>>>> >>>>>> -- Rob >>>>>> >>>>>> >>>>>> On 18 February 2014 20:41, Fraser Adams < >>>>>> [email protected] >>>>>> >>>>>> wrote: >>>>>>> >>>>>>> Hey Rob and Robbie, >>>>>> >>>>>> I've just quickly tried a rebuild of r1569501 following my previous >>>>>>> issue >>>>>>> about the cpp stuff failing to compile. I haven't built for a couple >>>>>>> of >>>>>>> weeks but I've just rebuilt the Java Broker and tried rebuilding the >>>>>>> QMF >>>>>>> stuff including the Java Broker plugin and got the errors below. It >>>>>>> looks >>>>>>> like there has been a change to one or other of the Broker Plugin >>>>>>> APIs. >>>>>>> Unfortunately I don't have a great deal of time to look at this at >>>>>>> the >>>>>>> moment 'cause it's half term here and I've got family stuff. Kind of >>>>>>> worried if something that's now broken has made it onto the the 0.28 >>>>>>> branch. I don't know when this got changed, but I think fairly >>>>>>> recently >>>>>>> (sometime over the last two or three weeks I guess). >>>>>>> >>>>>>> [javac] /home/fadams/qpid/qpid-trunk/ >>>>>>> qpid/tools/src/java/src/qpid- >>>>>>> broker-plugins-management-qmf2/java/org/apache/qpid/server/qmf2/ >>>>>>> QmfManagementAgent.java:234: >>>>>>> cannot find symbol >>>>>>> [javac] symbol : method getSubscriptions() >>>>>>> [javac] location: interface org.apache.qpid.server.model. >>>>>>> Session >>>>>>> [javac] if (session.getSubscriptions() != >>>>>>> null) >>>>>>> [javac] ^ >>>>>>> [javac] /home/fadams/qpid/qpid-trunk/ >>>>>>> qpid/tools/src/java/src/qpid- >>>>>>> broker-plugins-management-qmf2/java/org/apache/qpid/server/qmf2/ >>>>>>> QmfManagementAgent.java:236: >>>>>>> cannot find symbol >>>>>>> [javac] symbol : method getSubscriptions() >>>>>>> [javac] location: interface org.apache.qpid.server.model. >>>>>>> Session >>>>>>> [javac] for (Consumer subscription : >>>>>>> session.getSubscriptions()) >>>>>>> [javac] ^ >>>>>>> [javac] /home/fadams/qpid/qpid-trunk/ >>>>>>> qpid/tools/src/java/src/qpid- >>>>>>> broker-plugins-management-qmf2/java/org/apache/qpid/ >>>>>>> server/qmf2/agentdata/Broker.java:566: createQueue(java.util.Map< >>>>>>> java.lang.String,java.lang.Object>) in org.apache.qpid.server.model. >>>>>>> VirtualHost >>>>>>> cannot be applied to (java.lang.String,org.apache. >>>>>>> qpid.server.model.State, >>>>>>> boolean,boolean,org.apache.qpid.server.model. >>>>>>> LifetimePolicy,long,java.util.Map<java.lang.String,java. >>>>>>> lang.Object>) >>>>>>> [javac] Queue queue = >>>>>>> vhost.createQueue(nameParser.getQueueName(), State.ACTIVE, durable, >>>>>>> false, >>>>>>> [javac] ^ >>>>>>> [javac] /home/fadams/qpid/qpid-trunk/ >>>>>>> qpid/tools/src/java/src/qpid- >>>>>>> broker-plugins-management-qmf2/java/org/apache/qpid/ >>>>>>> server/qmf2/agentdata/Exchange.java:138: >>>>>>> cannot find symbol >>>>>>> [javac] symbol : variable AUTO_DELETE >>>>>>> [javac] location: class org.apache.qpid.server.model. >>>>>>> LifetimePolicy >>>>>>> [javac] boolean autoDelete = (lifetimePolicy == >>>>>>> LifetimePolicy.AUTO_DELETE) ? true : false; >>>>>>> [javac] ^ >>>>>>> [javac] /home/fadams/qpid/qpid-trunk/ >>>>>>> qpid/tools/src/java/src/qpid- >>>>>>> broker-plugins-management-qmf2/java/org/apache/qpid/ >>>>>>> server/qmf2/agentdata/Queue.java:144: cannot find symbol >>>>>>> [javac] symbol : variable AUTO_DELETE >>>>>>> [javac] location: class org.apache.qpid.server.model. >>>>>>> LifetimePolicy >>>>>>> [javac] boolean autoDelete = (lifetimePolicy == >>>>>>> LifetimePolicy.AUTO_DELETE) ? true : false; >>>>>>> [javac] ^ >>>>>>> [javac] 5 errors >>>>>>> >>>>>>> >>>>>>> ------------------------------------------------------------ >>>>>>> --------- >>>>>>> To unsubscribe, e-mail: [email protected] >>>>>>> For additional commands, e-mail: [email protected] >>>>>>> >>>>>>> >>>>>>> >>>>>>> ------------------------------------------------------------ >>>>>>> --------- >>>>>>> >>>>>> To unsubscribe, e-mail: [email protected] >>>>> For additional commands, e-mail: [email protected] >>>>> >>>>> >>>>> >>>>> --------------------------------------------------------------------- >>> To unsubscribe, e-mail: [email protected] >>> For additional commands, e-mail: [email protected] >>> >>> >>> > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > >
