Hello, I'm new to Talend and trying to configure it for high availability. (I hope I put this into the right forum.) I have two questions related: 1) If I want to cluster the TAC, I understand that they have to point to the same DB (or to replicated DBs, where the data replication is managed outside the TAC). However, what about the case when each TAC points to its own embedded H2 database? I think, I made this work initially, and the database replication was managed internally somehow, but I might be wrong. So, the question generally is: if each TAC instance points to its own embedded H2 database, will the DB replication between these two instances be made automatically without further configuration? Or do I really have to use same DB (of whatever type) or replicate the DBs manually? 2) When I want to create a virtual server, the documentation says I should use the Virtual server tab in TAC. However, I see only Job conductor, Execution plan and Servers in the Conductor menu. Maybe I don't have the right packages for high availability? Thanks Jan
After a little investigation, I found out the the virtual servers thing is a license issue. So the question is, is a TAC clustering enough to have also fail-over for job servers? When one machine with TAC+job server #1 is down, will the job be done on another machine with another job server #2 (supposing we have duplicated TAC) when the job is scheduled for the job server #1)? I guess I need virtual servers for this, right?
A little bit discussing with myself, however - the virtual server issue seems resolved to me. The answer is yes, I need virtual servers. However, I still have a question to the TAC clustering - will repeat it to be more clear. I would expect that for High Availability, I need only to set the quartz properties and then duplicate the tomcat servers + webapps. When the webapp is pointing to the embedded H2 database (each node to its own), I would expect the H2 DB to replicate automatically. Should it work like this?