As a PM/tech lead for WDQS, I want to know what are the different and best options are for moving WDQS off of Blazegraph, so that I can decide which option to move forward with in scaling up WDQS beyond Blazegraph's current capabilities.
Blazegraph is unmaintained and is reaching/has reached its limits for handling the scale of Wikidata and the volume/complexity of queries to it, causing a number of current problems and making future catastrophic failures probable. It is very likely/desirable that we update to a new backend to continue scaling service. In order to do so we should have a good understanding of what our different options are.
Previous research is captured in this sheet: https://docs.google.com/spreadsheets/d/1MXikljoSUVP77w7JKf9EXN40OB-ZkMqT8Y5b2NYVKbU/edit?usp=sharing , but it is outdated. We should re-examine still relevant options as well as new ones along relevant parameters to evaluate good options to replace Blazegraph.
- create list of criteria for evaluating candidates (probably don't want as many as the previous sheet had?)
- create list of candidates to evaluate (https://db-engines.com/en/ranking/graph+dbms might be helpful)
- a comparison spreadsheet of different graph backend options (similar to previous research)
- This paper may provide a good overview/survey of possible options: https://arxiv.org/pdf/2102.13027.pdf