Support for JackRabbit cluster

[Santosh Prabhu <sp...@savvion.com>]

Hi,

I am new to JackRabbit and looking forward to use JackRabbit in our
application.
We have a Work Flow management system which runs in clustered environment.
We cannot use JackRabbit in a client-server model because this creates a
single point of failiure. 
So we are trying to use JackRabbit as embedded in the JVM.
When the Repository is initialized it creates .lock file in the repository
home directory.
This file is used to prevent 2 clients using the same repository home
directory simultaneously.
I was just wondering if we could use database persistence and use 2 clients
pointing to the same database.

So I tried commenting out the lock file creation part in RepositoryImpl
class.
This way I was able to connect to the same repository from 2 clients.
But now I find that the updates made from one client are not visible to the
other.

As JackRabbit architecture says the changes made from 1 session would be
published to other active sessions. 
Is it possible that we can get the changes in sessions created on other
client programs?
Has someone tried out this kind of approach?

I serached the forums for related articles and found one thread discussing
the support for clustered environment. Is JackRabbit thinking of supporting
cluster in near future?

Thanks in advance for any coments.

Regards,

Santosh Prabhu
sprabhu@savvion.com
--
View this message in context: http://www.nabble.com/Support-for-JackRabbit-cluster-t1311529.html#a3494646
Sent from the Jackrabbit - Dev forum at Nabble.com.

Re: Support for JackRabbit cluster

[Chandresh Turakhia <ch...@bhartitelesoft.com>]

c-jdbc ( clustered jdbc - open source project ) should help a bit atleast

Chand
----- Original Message ----- 
From: "Santosh Prabhu" <sp...@savvion.com>
To: <ja...@incubator.apache.org>
Sent: Monday, March 20, 2006 6:51 AM
Subject: Support for JackRabbit cluster


>
> Hi,
>
> I am new to JackRabbit and looking forward to use JackRabbit in our
> application.
> We have a Work Flow management system which runs in clustered environment.
> We cannot use JackRabbit in a client-server model because this creates a
> single point of failiure.
> So we are trying to use JackRabbit as embedded in the JVM.
> When the Repository is initialized it creates .lock file in the repository
> home directory.
> This file is used to prevent 2 clients using the same repository home
> directory simultaneously.
> I was just wondering if we could use database persistence and use 2 
> clients
> pointing to the same database.
>
> So I tried commenting out the lock file creation part in RepositoryImpl
> class.
> This way I was able to connect to the same repository from 2 clients.
> But now I find that the updates made from one client are not visible to 
> the
> other.
>
> As JackRabbit architecture says the changes made from 1 session would be
> published to other active sessions.
> Is it possible that we can get the changes in sessions created on other
> client programs?
> Has someone tried out this kind of approach?
>
> I serached the forums for related articles and found one thread discussing
> the support for clustered environment. Is JackRabbit thinking of 
> supporting
> cluster in near future?
>
> Thanks in advance for any coments.
>
> Regards,
>
> Santosh Prabhu
> sprabhu@savvion.com
> --
> View this message in context: 
> http://www.nabble.com/Support-for-JackRabbit-cluster-t1311529.html#a3494646
> Sent from the Jackrabbit - Dev forum at Nabble.com.
>
>