I'm solving a problem that seems most appropriately handled by a graph database, so I wanted to get a graph database server up and running, and go from there. I'm a Python developer, so I was trying to get something running with the bulbs library, which seems mature and effective, based on the documentation.
Unfortunately, I haven't been able to find any monolithic guide that covers everything between bulbs and an actual graph database server, and my attempts to cobble together working versions have been hampered by a number of compatibility problems.
I feel like I might be missing something intrinsic to the design of these systems. I'm used to postgresql, MariaDB, and other systems which are a pretty simple two-layer model, bridged by a standard API. It seems like the Apache Tinkerpop stack should be what I want, but Rexster seems like it's a server, but not a storage backend, so I still need one of those? I'm a little confused, because Neo4j and Titan seem like they're also servers, in addition to storage backends, so I don't know why Rexster is necessary. Right now, I'm trying to get Neo4j to work with bulbs, but the Gremlin plugin is missing... I've spent more than a day trying to piece this software stack together, and I'm getting really close to just giving up and building a million mapping tables in an ORM.
Is there a monolithic installation guide that I can follow somewhere, or has anyone had experience getting this working in a sane amount of time? I can use any solution deployable on Fedora, Debian, or OpenBSD.
