4 ms·
Real-Time Databases Explained: Why Meteor, Rethink, Parse and Firebase Don't Scale
- code-is-code 9y agoMeteor has the op-log-driver which calculates query-result for changes instead of running the wohle query again. Same comes with rxdb where it's called query-change-detection.
- wolle 9y agoThat's true, but monitoring the stream of change events in a single machine (be it server or client) makes this machine your bottleneck for overall system throughput.
- btown 9y agoWritten format: https://medium.baqend.com/real-time-databases-explained-why-meteor-rethinkdb-parse-and-firebase-dont-scale-822ff87d2f87 https://medium.baqend.com/real-time-databases-explained-why-... And the paper for their more scalable alternative: https://www.cl.cam.ac.uk/~mks40/pubs/vldb_2017.pdf https://www.cl.cam.ac.uk/~mks40/pubs/vldb_2017.pdf The devil will absolutely be in the details, but the speaker/author identifies a real pain point that bolting on a replication-log tailing system to provide realtime capabilities results in numerous write bottlenecks unless the system is architected from scratch to avoid them.