jsr356-experts@websocket-spec.java.net

[jsr356-experts] Re: [jsr356-users] Re: Summary: Web Socket Sessions in Distributed Servers

From: Danny Coward <danny.coward_at_oracle.com>
Date: Fri, 07 Dec 2012 15:38:45 -0800

Hi Mark,

On 12/7/12 9:50 AM, Mark Thomas wrote:
> On 07/12/2012 02:12, Danny Coward wrote:
>> So I think that means websocket applications that are 'distributable'
>> should not
>> - use static variables or local filesystems to store application state
>> (use a database instead)
>> - put non Serializable data in websocket Sessions
>>
>> Let me know if you have more things you think the specification should
>> define for deployments on distributed servers.
> That implies some API changes. Session and MessageHandler will need to
> implement Serializable.
I think containers that are distributable will need to implement the
Session interface with a class that is serializable, and..yes I think
developers writing MessageHandlers that they would like to work in a
distributed container will need to make them Serializable.

Not all containers will be distributed, and not all distributed
containers will do failover. So I don't think we need to make everyone
do it.

> I'm guessing that RemoteEndpoint and
> ClientContainer can be reconstructed post fail-over.
>
> Do we want to add anything along the lines of
> HttpSessionActivationListener? My guess is that if it is required for
> HttpSEssion then we'll probably need it for the user properties map.
Yes be probably should.

- Danny
>
> Mark
>


-- 
<http://www.oracle.com> 	*Danny Coward *
Java EE
Oracle Corporation