Solr gets the query on some node. It then sends sub-requests out to one replica for each shard, perhaps on other nodes. You'd have to insure that the parameters get added before the subrequests are made. In the logs you'll see things like DISTRIB=false for sub-requests, so the first thing to check is whether your added params are on the subrequests. You'd have to alter the query itself here.
But I'd ask why put this on Solr? Why not do this at the client level? Not having control over the client is usually a good reason, but I thought I'd ask ;) FWIW, Erick On Wed, Dec 13, 2017 at 10:45 AM, Steve Pruitt <bpru...@opentext.com> wrote: > I am trying to work out how to share a parameter value in all the Solr nodes > in a cluster. > > I have two custom search components. One is defined in the first-components > list and the other is defined in the last-components list. Ideally, the > first-component creates the parameter value and sets so the same component > for the other nodes can find it. > > The request SolrParams are good for sharing within the same node, but I don't > know if the request defined by the aggregator for the other shards is exposed. > > Hopefully this makes sense. > > I also have noticed that the custom components are executed twice for each > query. Not sure why. > > Thanks in advance. > > -S >
