We benchmarked Pulsar sync times. Verdict? The algorithm scales and performs beautifully!
August 29, 2016
One of the critical concerns for anyone dealing with Offline Synchronization is scaling. How much data can you download and how fast? The scope and complexity of the problem quickly escalates if you are dealing with mobile devices as the offline store. It is very typical for our enterprise customers to have large amounts of data and they expect the users to have access to this without interruption no matter whether they are online or offline. Our internal tests proved that the algorithm scales really well but we also had a chance to measure real sync times from some of our customers with real world scenarios (people with spotty 3G connection to robust WiFi times). We decided to publish some of those numbers since they are very very impressive!
Note: Please know that these are real customer data sets and sync times. We can arrange a reference call if you would like to talk to them but we can’t publish the names of the clients here.
Initial Sync Times
This is the first time user is logging into the app, so they are downloading all of the data for the first time.
- 3.1 Million Records downloaded in 47 Minutes (3G connection, iPad Air2).
- 1.2 Million Records downloaded in 24 Minutes (WiFi connection, iPad Air2).
- 900K Records downloaded in 19 Minutes (WiFi connection, iPad3).
- 100K Records downloaded in 6 Minutes (3G connection, iPad3).
Catch-up Sync Times
This is a delta sync, meaning the user logged into the app after a successful initial sync and the app syncs data that was changed since the last successful sync.
- 72K Records downloaded in 3.5 Minutes (3G connection, iPad Air2).
- 32K Records downloaded in 5 Minutes (Spotty WiFi connection, iPad Air2).
- 9K Records downloaded in 0 Minutes 52 Seconds (Good WiFi connection, iPad Air2).
Catch-up Sync With Schema Change Times
This happens when the object schema on the backend changed. Example, the admin added a couple of fields or took away a field from the object field set and the app has to handle that situation and make sure the local DB is in sync with those changes.
- 22K Records downloaded with handling of schema change and reindexing the DB. 7 Minutes (Good 3G connection, iPad Air2).
Conclusion
To conclude, Pulsar handles massive datasets really well and our customers can speak to that fact. So, please reach out if you have any questions!
Follow Us for More Exciting Updates!
As we continue to innovate and enhance our software, we invite you to stay connected and be the first to know about future releases, product tips, and much more! Follow our blog to access in-depth articles, engaging tutorials, and expert perspectives that empower you to make the most of our software’s capabilities.
Connect with us on our social media channels to become part of a vibrant community of like-minded individuals. Share your experiences, provide feedback, and join the conversation as we collectively shape the future of software innovation.
Share on: