Hi Asela,

Thanks for the information. I implemented above and tested for a cluster
with two nodes and it worked as expected.

On Tue, Sep 9, 2014 at 12:47 PM, Asela Pathberiya <[email protected]> wrote:

> Hi Hasintha,
>
> Sorry for late response....
>
> On Wed, Sep 3, 2014 at 6:37 PM, Hasintha Indrajee <[email protected]>
> wrote:
> > Hi Asela,
> >
> > I am working on the issue [1]. Can you please elaborate bit on the cache
> > implementation of PIP side in XACML. As far as I noticed, as a solution
> for
> > the bug part of the issue, invalidating Attribute cache and Decision
> cache
> > is sufficient when an update occurs to user information (when firing user
> > management listener).
> >
> > Are we using hazelcast for caching in XACML ? If not do we need to send
> > cluster messages on the event of cache invalidation ?. Has this been
> already
> > implemented in any of the caches in XACML ?
>
> Yes.. Entitlement component is using the Hazelcast...  Therefore you
> do not need to send cluster messages...
>
> >
> > What are the usages (differences) of DecisionCache and
> > DecisionInvalidationCache ?
>
> Decision cache keeps the XACML request and response and it is a local
> cache.   DecisionInvalidationCache cache is used to invalidate the
> local caches in the cluster node when any update or API call is done.
>
> Thanks,
> Asela.
>
> >
> >
> > [1] https://wso2.org/jira/browse/IDENTITY-2567
> > [2] https://redmine.wso2.com/issues/2901 - redmine issue
>
>
>
> --
> Thanks & Regards,
> Asela
>
> ATL
> Mobile : +94 777 625 933
>
_______________________________________________
Dev mailing list
[email protected]
http://wso2.org/cgi-bin/mailman/listinfo/dev

Reply via email to