On Fri, Jul 20, 2012 at 2:30 PM, Subash Chaturanga <[email protected]> wrote:

>
>
> On Fri, Jul 20, 2012 at 2:17 PM, Tharindu Mathew <[email protected]>wrote:
>
>> Alright, so it worked earlier cus caching was not there at the registry
>> API level?
>
>
> Yes Tharindu you are quite right.
>
> Earlier (may be in our internship period time ;-) ) greg did  not have
> registry API level caching and now registry calls intercepted by a registry
> called CacheBackedRegistry. And those days as there is no cache, no need
> replicated caching enabled as both ends directly deals with the DBs.
>
How is the performance with replicated caching in a mounted scenario like
this compared to the non cached mounted scenario? I hope not having a cache
with jdbc mounting was not faster, although it could have been a conscious
decision made during your intern time for faster performance ;)

>
>>
>> On Fri, Jul 20, 2012 at 2:14 PM, Subash Chaturanga <[email protected]>wrote:
>>
>>>
>>>
>>> On Fri, Jul 20, 2012 at 1:19 PM, Tharindu Mathew <[email protected]>wrote:
>>>
>>>> There is no ESB cluster here. Only an ESB connecting to G-Reg. I
>>>> believe for this scenario, replicated caching is not needed?
>>>
>>> Yes, I also didn't meant ESB cluster(may be I have not communicated it
>>> ;-) ). I am also talking about a typical mounted scenario, i.e 1 esb, 1
>>> greg,  esb gov mounted to greg gov and etc. And AFAIU the scenario I
>>> explained is valid for such mounts.
>>>
>>>
>>>
>>>>
>>>>
>>>> On Fri, Jul 20, 2012 at 12:36 PM, Subash Chaturanga <[email protected]>wrote:
>>>>
>>>>>
>>>>>
>>>>> On Fri, Jul 20, 2012 at 10:36 AM, Tharindu Mathew 
>>>>> <[email protected]>wrote:
>>>>>
>>>>>> Subash,
>>>>>>
>>>>>> Can I know why this is needed? This was not required earlier so I'd
>>>>>> like to understand why it's needed now?
>>>>>>
>>>>>
>>>>> Hi Tharidu,
>>>>> I hope you meant why caching a must for jdbc mounting. I am not sure
>>>>> why this didn't wanted earlier.
>>>>>
>>>>> In caching of carbon clusters, each carbon instance has its own cache.
>>>>> Assume replicated caching not enabled.
>>>>>
>>>>> So suppose a two node cluster which has jdbc mount. When do a update
>>>>> on one carbon instance and it clears the cache and persist the
>>>>> change(delete/put) to the DB. The problem comes when some other node do a
>>>>> *get* from the same earlier changed location. Because its cache not
>>>>> updated. Unless you do a put on the same location and clears the cache you
>>>>> will not see the changes. If replicated caching enabled, infinispan does
>>>>> the sync between two caches.
>>>>>
>>>>>
>>>>> Hi Charitha,
>>>>> In fact this is a MUST for atom/ws as well IF you need 100%
>>>>> consistency across the cluster. Because there are some edge cases.
>>>>> It works fine until you do changes from the mounted side. If you do
>>>>> some changes on the other side(master), the changes not get reflected. I
>>>>> hope you got what I meant.
>>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>
>>>>>> On Fri, Jul 20, 2012 at 9:49 AM, Charitha Kankanamge <
>>>>>> [email protected]> wrote:
>>>>>>
>>>>>>> Thanks Subash, this is something we have not documented anywhere. We
>>>>>>> should update all docs, blogs with this information.
>>>>>>>
>>>>>>>
>>>>>>> On Friday, July 20, 2012, Subash Chaturanga wrote:
>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Thu, Jul 19, 2012 at 11:48 PM, Charitha Kankanamge <
>>>>>>>> [email protected]> wrote:
>>>>>>>>
>>>>>>>>> Folks,
>>>>>>>>> Should replicated caching be enabled with registry mounting?
>>>>>>>>> Suppose /_system/governance space of an ESB instance is mounted (jdbc 
>>>>>>>>> or
>>>>>>>>> ws) to the same collection of a central G-reg. In that case, should we
>>>>>>>>> enable replicated caching in both ESB and G-reg instances?
>>>>>>>>> Please confirm.
>>>>>>>>>
>>>>>>>> Hi Charitha,
>>>>>>>> Yes, for JDBC mounting it is a MUST. But AFAIK, for atom/ws
>>>>>>>> mounting it should be fine without replicated caching enabled.
>>>>>>>>
>>>>>>>>
>>>>>>>>>
>>>>>>>>> Thanks!
>>>>>>>>> Charitha
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>>> Dev mailing list
>>>>>>>>> [email protected]
>>>>>>>>> http://wso2.org/cgi-bin/mailman/listinfo/dev
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>>
>>>>>>>> Subash Chaturanga
>>>>>>>> Software Engineer
>>>>>>>> WSO2 Inc. http://wso2.com
>>>>>>>>
>>>>>>>> email - [email protected]
>>>>>>>> phone - 077 2225922
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Charitha Kankanamge
>>>>>>> cell: +94 718 359 265
>>>>>>> blog: http://charithaka.blogspot.com <http://wso2.com>
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Dev mailing list
>>>>>>> [email protected]
>>>>>>> http://wso2.org/cgi-bin/mailman/listinfo/dev
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Regards,
>>>>>>
>>>>>> Tharindu
>>>>>>
>>>>>> blog: http://mackiemathew.com/
>>>>>> M: +94777759908
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>>
>>>>> Subash Chaturanga
>>>>> Software Engineer
>>>>> WSO2 Inc. http://wso2.com
>>>>>
>>>>> email - [email protected]
>>>>> phone - 077 2225922
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Regards,
>>>>
>>>> Tharindu
>>>>
>>>> blog: http://mackiemathew.com/
>>>> M: +94777759908
>>>>
>>>>
>>>
>>>
>>> --
>>>
>>> Subash Chaturanga
>>> Software Engineer
>>> WSO2 Inc. http://wso2.com
>>>
>>> email - [email protected]
>>> phone - 077 2225922
>>>
>>>
>>
>>
>> --
>> Regards,
>>
>> Tharindu
>>
>> blog: http://mackiemathew.com/
>> M: +94777759908
>>
>>
>
>
> --
>
> Subash Chaturanga
> Software Engineer
> WSO2 Inc. http://wso2.com
>
> email - [email protected]
> phone - 077 2225922
>
>


-- 
Regards,

Tharindu

blog: http://mackiemathew.com/
M: +94777759908
_______________________________________________
Dev mailing list
[email protected]
http://wso2.org/cgi-bin/mailman/listinfo/dev

Reply via email to