I understand that this is a strange question, but for reasons that defy logic my users want the following setup and after much google-ing and testing I can't figure out if it's possible:
a) There are 5 users, each at their own workstation.
b) There is one set of source data to index on a file share that all workstations have access to.
c) The users must run local elastic instances on their workstations - there can be no centralized Elastic server, and no Elastic instance running on the file server.
d) Any one of the workstations can initiate indexing the source data on the file server.
e) The index data itself must reside on the file server. Index data must not be stored locally on the workstation(s).
f) After indexing, all 5 workstation-local instances must be able to query the newly created index.
g) Not required, but also requested: All of the index is contained in a folder that can be moved to arbitrary locations (may require a config change to be able to connect to after moving).
I thought that this maybe could be accomplished by setting the "path.data:" on each of the workstation instances to the same directory on the shared file server so that when one workstation instance does the indexing, the other workstation instances could see and query it. This also seems like it could cause all kinda of conflicts.
Is there any way to accomplish what I've described above?
Note that for this application, there is no need for replication, and the entire index can be in one shard. In case any of that makes things easier.