synapse-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Supun Kamburugamuva <supu...@gmail.com>
Subject Re: Deployer implementation for synapse
Date Tue, 25 Aug 2009 07:23:00 GMT
Hi,
If the user can turn off hot update then this won't be a performance issue.
I believe this is the way you are implementing it. So if a user wants more
flexibility they have to sacrifice performance. If they
want performance they shouldn't go for things like hot update.

Thanks,
Supun..

On Tue, Aug 25, 2009 at 12:11 AM, Ruwan Linton <ruwan.linton@gmail.com>wrote:

> Yes Supun, the SynapseConfiguration edits are not synchronized for the
> moment. We will any way need to make them synchronized with enabling the hot
> update.
>
> I am working on that. I prefer to implement with our own deployer
> implementation, because it seems we do not have the flexibility that is
> required with the axis2 deployer architecture.
>
> Thanks,
> Ruwan
>
>
> On Tue, Aug 25, 2009 at 11:52 AM, Supun Kamburugamuva <supun06@gmail.com>wrote:
>
>>
>>
>> On Mon, Aug 24, 2009 at 3:57 AM, Ruwan Linton <ruwan.linton@gmail.com>wrote:
>>
>>> Folks,
>>>
>>> I was trying to implement the hot deployment and hot update of the
>>> synapse artifacts as per my proposal the to the list sometime back using the
>>> axis2 deployers. But due to some restrictions and the architecture of the
>>> axis2 deployers we cannot use that for synapse artifact deployment.
>>>
>>> The axis2 deployer hot update functionality has been implemented to call
>>> the unDeploy() method and then call the deploy method on that artifact. This
>>> causes an issue for syanpse since synapse has some restrictions on a
>>> absolute minimal configuration. We cannot let the user undeploy the main and
>>> fault sequences. But acording to the way the axis2 hot update has been
>>> implemented we do not have a means of differentiating the hot update
>>> undeploy call from the actual undeploy call. Actual undeploy call should
>>> prevent undeploying the main and fault sequences while it should be possible
>>> to undeploy those for hot updating.
>>>
>>> Also please note that with this approach there is a donwtime of the
>>> artifacts when doing the hot update which has to be prevented some how.
>>>
>>> Given the above facts it seems like we will have to go with a synapse
>>> implementation of the deployers.
>>>
>>
>> If we go for a deployment like this, the implementation may require quite
>> a lot of synchronizations. I'm not sure how the implementation is going to
>> be. But I just want to bring this up so we can discuss if there is an issue.
>>
>> Thanks,
>> Supun..
>>
>>
>>>
>>> WDYT??
>>>
>>> Thanks,
>>> Ruwan
>>>
>>> --
>>> Ruwan Linton
>>> Technical Lead & Product Manager; WSO2 ESB; http://wso2.org/esb
>>> WSO2 Inc.; http://wso2.org
>>> email: ruwan@wso2.com; cell: +94 77 341 3097
>>> blog: http://ruwansblog.blogspot.com
>>>
>>
>>
>>
>> --
>> Software Engineer, WSO2 Inc
>> http://wso2.org
>> supunk.blogspot.com
>>
>>
>>
>
>
> --
> Ruwan Linton
> Technical Lead & Product Manager; WSO2 ESB; http://wso2.org/esb
> WSO2 Inc.; http://wso2.org
> email: ruwan@wso2.com; cell: +94 77 341 3097
> blog: http://ruwansblog.blogspot.com
>



-- 
Software Engineer, WSO2 Inc
http://wso2.org
supunk.blogspot.com

Mime
View raw message