Recently, we announced the release of the latest version of our industry-leading Real User Monitoring solution, mPulse. The updated release, mPulse 54, features a range of new capabilities to ensure the solution is always at its best.
One of the most exciting changes in mPulse 55 is the ability to leverage several DataStores within instances separate from the DataService instance. This provides added elasticity at the data layer level, as each DataStore runs on its own PostgreSQL database.
This is important for a number of reasons, including the fact that users can now access more storage space for aggregate beacons being collected during the RUM process. As each individual DataStore reaches capacity, it can be split up at runtime to allow for a fraction of the data to be stored there.
Multiple DataStores also enables data to be partitioned in new ways. Currently, mPulse separates data according to its date and domain. With the new release, user data is striped across all DataStores being utilized, which considerably enhances the performance of large queries. In our internal tests, we saw a marked improvement with this capability in place, allowing for one client's dashboard – which included 1.5 billion beacons collected over 30 days – to reduce load time from 4 minutes to 15 seconds. Now that's what we call fast.
Other additions and improvements
In addition, the new mPulse update features added entry and exit queues for each Collector and Consolidator. This ensures that even when there is a serious spike in beacon volume, users have adequate CPU or memory resources to complete their tasks.
SOASTA is always looking for ways to ensure that our best-in-class systems remain leading solutions in the industry, and this represents just a small part of those efforts.
For more information on mPulse, contact SOASTA today.
About the Author