dbStripe adds Striping, Replication, and/or HA to any existing SQL Server database. Its purpose is to improve performance by splitting data over multiple processor cores, hard drives, and servers. dbStripe is designed for large amounts of data (terabytes or higher), but can also improve performance for small databases that have high query traffic.
dbStripe is extensible, centrally managed, failsafe, and has no bottlenecks. It resides within the database, and is completely transparent to applications and existing database scripts. dbStripe allows data tables to be split across SQL databases on a single server or multiple servers, including geographically dispersed ones. Replication duplicates data to another server. Any of these features can be configured for one or more tables, in any combination, requiring no application changes.
Features of dbStripe
dbStripe runs on Microsoft SQL Server only, and at present is in Beta stage.
- Supports geographically dispersed super-clusters
- Up to 32767 SQL Instances per cluster
- No limit on number of clusters
- Not limited to key ranges like Partitioning
- No unique key field needed
- Multi read and write instances as needed
- Also provides replication and HA for any Edition
Reasons to buy dbStripe
Here are the major reasons to consider the product:
- Improve performance by striping/replication
- Replicated tables can have different indexes, further improving performance
- Provide failure-redundancy via replication
- Reduce/eliminate need to upgrade to SQL Server Enterprise Edition
- Transparent to application programs, so no reprogramming
- Works with packaged products having no source code