S3 Lucene Implementation?

Feb 16, 2011 at 1:15 PM

Just curious if anyone has any thoughts on modifying the existing Lucene full text indexing functionality in Simple Savant (excellent implementation BTW) to store the indexes in S3 as opposed to the local file system.

I don't actually even know if it is possible to host the index files in S3 (locking, etc...) but it just seems like it would be a good solution to distributing the index between multiple applications.

Thanks,

Hal

Coordinator
Feb 16, 2011 at 1:37 PM

Storing indexes in S3 might be possible, but would likely be quite difficult to implement and perform poorly since s3 doesn't support random access. Storing indexes on a shared network drive would be much easier. I believe this is possible with EC2 instances, but I'm not sure of the steps to set it up or if third-party software is required.

Perhaps someone else has experience with sharing Lucene indexes on EC2 and can comment...