I’m working with Titan 0.4.4 and Cassandra 1.2.13 as my storage backend. My graph already has a bunch of vertices and edges. I just added a new type of vertex and I want to create an index on its ID. Is there a way to do this without having to start from scratch? I really don’t want to wipe out all my existing data and rebuild everything. I know it works if I clear the backend and recreate all the indices, but I’m hoping there’s a way to avoid that. Has anyone figured out how to add an index to an existing Titan database without losing data? Any tips or suggestions would be super helpful!
Adding an index to an existing Titan 0.4.4 database without data loss is indeed challenging. In my experience, the safest approach is to perform a full data export, recreate the schema with the new index, and then reimport the data. This method preserves your existing information while allowing for schema modifications. However, it can be time-consuming for large datasets. Alternatively, you might consider upgrading to a newer version of Titan or its successor, JanusGraph, which offer more flexible indexing options. If upgrading isn’t feasible, exploring third-party tools for graph database migration could provide a solution tailored to your specific needs.
hey max, i’ve been there before! have u tried using the TitanManagement API? it might let u add the index without wiping everything. just be super careful and backup ur data first. also, whats the size of ur graph? that could affect how long itll take. good luck!
hey there! im curious, have u considered using a manual indexing process? it might be a bit tedious, but could work without data loss. how big is ur dataset? maybe we could brainstorm some creative workarounds? whats the specific property youre trying to index?