Hi Bertrand,

The intended patch for SLING-3309 is generic, meaning that it just allows 
customization of the HTTP payload (headers, body, maybe method and url?).  That 
can be used for cache flushing but also for other things like backing up 
packages to a WebDav server for example.

Marius


-----Original Message-----
From: Bertrand Delacretaz [mailto:[email protected]] 
Sent: Wednesday, January 08, 2014 11:31 AM
To: dev
Subject: Re: [replication] Whiteboard pattern for replication events?

Hi Tommaso,

On Wed, Jan 8, 2014 at 10:17 AM, Tommaso Teofili <[email protected]> 
wrote:
> ...The problem I see here, as far as I understood eventing, is related 
> to the fact that events are not guaranteed to be "consumed" in the 
> exact same order they were "generated" and I think this would be a 
> problem for flushing a cache consistently or am I missing something?...

http://felix.apache.org/site/apache-felix-event-admin.html says "events from 
the same thread are sent in the order they are received (this is according to 
the spec)" - I'm not an expert on that but if that's correct you just need to 
send those OSGi events from a single thread to have a consistent ordering.

If this works I'd say the SLING-3309 is better handled outside of Sling, by 
listening to these replication events and making the corresponding HTTP 
requests, unless you think it's a simple change to the Sling replication code.

-Bertrand

Reply via email to