I'm using MySQL to keep my parent-child relation. The information is targeted in a single table which is really elegant by design.

I have no problem quering nevertheless the table is continuing to grow in dimensions significantly. I must shard or apply certain strategies to enhance the performance of my queries (and joins). Wouldso would I actually do it?

Any pointers could be appreciated.

Thanks, R

There's lack of knowledge within this publish to provide you with a great answer. The very first factor we will have to know is exactly what the table structure is. Have you got non-clustered indexes defined or perhaps is it really one clustered index.

The application that utilizes this database, exactly what does it use the information?

Additionally, looking for say a title area, or perhaps a specific date? If that's the case this can be candidates to have an index type.

The main factor to take care of though is indexes.

I believe you'll need more complex help in case your method is growing that large. dbShards is a great starting point. They can help you design a sharding technique for your computer data and may easily shard any tables that require it. In case your one table gets too big, sharding isn't a bad choice. I'd take a look at dbShards. Might work nicely with this situation.