jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Thomas Mueller <muel...@adobe.com>
Subject Re: Help regarding Jackrabbit clustering:
Date Thu, 09 Jun 2011 12:51:55 GMT
Hi,

Please note I don't blame you for making a mistake. The Jackrabbit
configuration is _really_ painfully complicated. It's extremely easy to
make such mistakes.


> Only the content goes to the one database

No. The FileSystem is also in the database according to the config you
sent.

>We are not using shared filesystem for workspace and repository.

According to the configuration you sent, this is not correct. You _are_
using a shared database file system:

<FileSystem class="org.apache.jackrabbit.core.fs.db.DbFileSystem">
  <param name="driver" value="org.postgresql.Driver"/>
  <param name="url" value="jdbc:postgresql://10.128.16.60:5432/picoemsdb"/>
  <param name="user" value="postgres"/>
  <param name="password" value="postgres"/>
  <param name="schema" value="postgresql"/>
  <param name="schemaObjectPrefix" value="jcr_workspaceTable_fs_"/>
</FileSystem>


Both cluster nodes use the same storage (and also all workspaces use the
same config, which is yet another problem). You should use something like:

<param name="url" 
value="jdbc:postgresql://10.128.16.60:5432/picoemsdbNode1"/>
<param name="schemaObjectPrefix" value="jcr_${wsp.name}_fs_"/>
and
<param name="url" 
value="jdbc:postgresql://10.128.16.60:5432/picoemsdbNode2"/>
<param name="schemaObjectPrefix" value="jcr_${wsp.name}_fs_"/>

I don't actually know if ${wsp.name} is supported or if you need
${wsp.home}.

After storing, you need to wait up to 5 seconds before you read the data
on the other cluster node (due to the sync delay).

Regards,
Thomas





Mime
View raw message