On Tue, May 4, 2010 at 10:57 PM, indika kumara <indika.k...@gmail.com> wrote:
> Ruwan > > >> but JMX management and error notifications like suspensions and so forth >> cannot be controlled if those endpoints are inlined in, say send mediator. >> The JMX management and all that are only possible for declared endpoint >> > > I am not sure this ... I have not looked the endpoint code for a long > time..But as remind , It should just need a name and does not depend on > in-lined / declared > > >> The only enforcement that we can do to get rid of all that is to disable >> the ability to have inlined endpoints, but I am strongly -1 on this >> approach, since it reduces a usability a lot. >> >> Also, please note that we do not sacrifice correctness by any means hear, >> if we do this enforcement, we loose the usability and if we keep this as it >> is then again what we will loose is the usability, but the usability >> reduction on the first approach is huge with compared to doing this change. >> >> Hence, I am -1 on enforcing a name for inlined endpoints or disabling the >> ability to have inlined endpoints. >> >> Thanks, >> Ruwan >> >> > I also do not like to remove in-lined endpoints.... but ....making > endpoint name mandatory brings advantages than disadvantages to a user even > he may not be aware at first time the importance of a meaningful name. > Well other than the disadvantage of huge usability degradation, I don't actually see any advantage in enforcing names for in-lined endpoints. Endpoint management facilities cannot be considered as an advantage. If the user wants to manage endpoints then he can give them meaningful names. That option is always there and it is the best practice. But Synapse should not enforce the user to do so. Note that this will also bring in some migration problems. I don't think we can write a migration tool which can get a config with anonymous endpoints and assign them meaningful names. So I too have to put my -1 on this. Thanks, Hiranya On Wed, May 5, 2010 at 6:48 AM, indika kumara <indika.k...@gmail.com> wrote: > > However, the clustering is not. It is an internal operation. >> >> > Sorry, I was wrong by.. The clustering is needed careful effort from > user. Unique names are mandatory and same names should be in all synapse > instances of the cluster. This cannot be done with the random name > generation as user do not know what are the generated names. He has to > manually put same names in each synapse instance. Therefore, with the random > name generation , the clustering will not work ... so In synapse , current > endpoint operation is incorrect .... > > > > > -- Hiranya Jayathilaka Software Engineer; WSO2 Inc.; http://wso2.org E-mail: hira...@wso2.com; Mobile: +94 77 633 3491 Blog: http://techfeast-hiranya.blogspot.com
