On 9/12/2017 11:27 AM, Shashank Pedamallu wrote: > I wanted to know how does Solr pick up cores on startup. Basically, what I > would like to try is Read cores created from one Solr instance from another > Solr instance. i.e., > > * I will have 2 Solr Instances SOLR1, SOLR2. Only SOLR1 is started. > * I’m creating a core (Core1) using SOLR1. > * After filling it to some capacity, I unload this core without deleting > the data. > * I would now start SOLR2 and would like to point SOLR2 to Core1. > Can someone please share me the details on how I can achieve this?
In Solr 4.4, the core discovery process became usable. In 4.x versions starting with 4.4, core discovery was optional, but recommended. https://wiki.apache.org/solr/Core%20Discovery%20%284.4%20and%20beyond%29 In Solr 5.0, the old solr.xml way of specifying cores was removed, and now Solr only supports core discovery. Through the CoreAdmin API, you can add Solr cores that are positioned anywhere on the filesystem, but if you want them to be found on startup, they must all be located in one place, typically underneath the solr home directory. You can use filesystem tricks like symlinks to have the true locations be variable. It would be a very bad idea to try and have two Solr instances pointing at the same solr home, unless you could be absolutely certain that they would never run at the same time. Instead, you should move or copy a core from one instance's solr home to that of another instance to transfer it. Thanks, Shawn
