Clustering II #3
Conversation
These changes will allow Jentrata to run as a cluster. The cluster can be build up out of several Jentrata instances on top of one Jentrata database. In this version only Postgresql is supported, but changes in the appropiate property files will allow to run it on top of other DBMSes. After installing this implementation make sure it will look at the central postgresql database. This can be set in the property file: ./plugins/hk.hku.cecid.ebms/conf/hk/hku/cecid/ebms/spa/conf/ebms.module. xml under the home directory of Jentrata. You need to change the line: <parameter name="url" value="jdbc:postgresql://localhost:5432/ebms" /> by replacing 'localhost' into the address of the central postgresql database. On other important setting is to make sure the 'hostname' can be translated into an ip address that's unique for the Jentrata server.
|
Jentrata » jentrata-msh #26 SUCCESS |
Additional code for the Cluster II. This code will audit the central database for hosts which are no longer available in the cluster but left messages unprocessed. This audit is part of each host within the cluster. The audit code itself is aware of the other hosts within the cluster. This is done via a newly created table called 'cluster'. In the cluster table each host will be represented by its ip-address, status (active, inactive, auditor) and a timestamp telling each host when it is allowed to run their audit.
|
Jentrata » jentrata-msh #27 FAILURE |
Previous commit did remove some code that which might have resulted in build errors, although no build errors were found in my environment. Hopefully these patches will fix it. To be complete this is the end of my my build output: [INFO] ------------------------------------------------------------------------ [INFO] Reactor Summary: [INFO] [INFO] Jentrata MSH: Parent ............................... SUCCESS [ 3.807 s] [INFO] Jentrata MSH: Ext .................................. SUCCESS [ 0.091 s] [INFO] Ext: EbXML ......................................... SUCCESS [ 7.130 s] [INFO] Ext: EbXML CPA ..................................... SUCCESS [ 3.386 s] [INFO] Jentrata: Commons .................................. SUCCESS [ 11.039 s] [INFO] Ext: Commons Extensions ............................ SUCCESS [ 0.944 s] [INFO] Core: Parent ....................................... SUCCESS [ 0.104 s] [INFO] Core: Corvus Core .................................. SUCCESS [ 1.140 s] [INFO] Jentrata: Plugins .................................. SUCCESS [ 0.028 s] [INFO] Plugins: Corvus AS2 ................................ SUCCESS [ 3.010 s] [INFO] Ext: Commons Test .................................. SUCCESS [ 1.287 s] [INFO] Core: Corvus Web ................................... SUCCESS [ 2.390 s] [INFO] Plugins: Corvus Main ............................... SUCCESS [ 1.142 s] [INFO] Plugins: Corvus Admin .............................. SUCCESS [ 1.437 s] [INFO] Plugins: Corvus JMS ................................ SUCCESS [ 16.368 s] [INFO] Plugins: Corvus EbMS ............................... SUCCESS [ 14.499 s] [INFO] Plugins: Corvus Main Admin ......................... SUCCESS [ 3.122 s] [INFO] Plugins: Corvus EbMS Admin ......................... SUCCESS [ 5.696 s] [INFO] Plugins: Corvus AS2 Admin .......................... SUCCESS [ 4.195 s] [INFO] Jentrata: Clients .................................. SUCCESS [ 0.046 s] [INFO] Clients: Corvus WSClient ........................... SUCCESS [ 49.458 s] [INFO] Jentrata MSH: Distribution ......................... SUCCESS [ 21.548 s] [INFO] ------------------------------------------------------------------------ [INFO] BUILD SUCCESS [INFO] ------------------------------------------------------------------------ [INFO] Total time: 02:33 min [INFO] Finished at: 2014-12-03T09:53:13+01:00 [INFO] Final Memory: 137M/330M [INFO] ------------------------------------------------------------------------
|
Jentrata » jentrata-msh #28 UNSTABLE |
|
Hi Hans (@cavycorp ) This pull request looks good. Will look into why the buildhive build failed. We do get intermittent build fails from time to time if that's the case I will merge this in. It might be worth while pushing out a release version with this new feature. cheers |
Remove the copy of the Duplicate Elimination setting from a message into the Acknowledgement service message. Since there is no need for this setting in any service message and there are implementations that can't handle this.
Update Acknowledgement Message
|
Jentrata » jentrata-msh #31 SUCCESS |
Revert "Update Acknowledgement Message"
|
Jentrata » jentrata-msh #32 UNSTABLE |
|
Jentrata » jentrata-msh #38 SUCCESS |
|
Hi @cavycorp as part of this pull request are you able to update the sql and ebms dao configs for the other DB types to support the clustering changes. |
|
Yes I think I can, let me look into that and I will get back to you with additional changes.
|
|
Jentrata » jentrata-msh #39 SUCCESS |
|
Jentrata » jentrata-msh #40 SUCCESS |
Add Oracle SQL en MySQL support in the DAO for the Cluster Audit
|
Jentrata » jentrata-msh #41 SUCCESS |
|
Oeps added changes with another account (rinis-dev) but the changes are valid. These should complete the Oracle SQL and MySQL DAO and Config changes. |
These changes will allow Jentrata to run as a cluster. The cluster can
be build up out of several Jentrata instances on top of one Jentrata
database. In this version only Postgresql is supported, but changes in
the appropiate property files will allow to run it on top of other
DBMSes.
After installing this implementation make sure it will look at the
central postgresql database. This can be set in the property file:
./plugins/hk.hku.cecid.ebms/conf/hk/hku/cecid/ebms/spa/conf/ebms.module.
xml under the home directory of Jentrata. You need to change the line:
by replacing 'localhost' into the address of the central postgresql
database.
Another important setting is to make sure the 'hostname' can be
translated into an ip address that's unique for the Jentrata server.