The released version works on a single machine. TupleFlow was originally built to work with Sun's Grid Engine to distribute computation to many machines. Since not many people have Grid Engine installed, we've decided not to release that part of the code. We hope to have the SSHStageExecutor class working soon so that TupleFlow can run across many machines without needing an external scheduler.
| [top] |
Galago is designed to be an experimental tool. Because of that, both the indexer and the retrieval engine are built from small parts that can be replaced with alternatives at run time. We hope that Galago makes it possible for researchers to build small custom indexing or retrieval components and share them with each other.
If you are looking for a search product and not a toolkit, other systems will suit your needs better.
| [top] |