Hi All, I hope this email finds there are still some people working on and discussing Jackrabbit 2 🙂
It’s very possible I am barking up a wrong tree here but I am trying to figure out how to properly do a hot backup of the Jackrabbit 2 Lucene indexes from a running cluster node. I started by reading the Lucene book which covers a single index but it seems Jackrabbit has its own index definition that makes things more complicated. From what I can tell, there is a combination of a volatile index alongside several persistent indexes per workspace. It’s not clear how to apply the “from the book” hot backup solution to Jackrabbit. I’ve also look at things like grabbing the current cluster local revision number, making a zip file and then doing some cleanup when starting a new node. Not that hard to program but very unclear if this guarantees a usable index. Also tried looking at things like somehow pausing a cluster node from processing updates while doing the backup but thus far very unclear if that’s possible. Maybe I’m barking up several of the wrong trees. Does anyone have guidance here? Is it even possible or is the only way to guarantee a consistent index to shutdown Jackrabbit? Tom
