In general yes, Parse Server is scalable and I am aware of deployments that manage at least millions of users, serving hundreds of thousands of DAUs.
We always aim for Parse Server to be versatile to cover as many use cases as possible. However, it depends on what your application does and the features and performance it requires. Parse Server - like any platform - comes with some specific characteristics. For example, it is based on MongoDB*, which comes with its own conditions.
Parse Server on the other hand is open source and allows you to have your own fork tailored to your needs and create your own plug-ins and adapters. We also offer a paid contributors program, so if there are specific features that you require, you could contribute some of your revenue back into the program and request specific features, which may then be added (depending on some considerations) and made available to the community.
The most significant shortcomings currently are:
-
The current release approach of manual releases in undefined release cycles while loosely following semver. We are currently preparing to move to a new release and versioning system to publish releases with vulnerability fixes faster. Currently we have months between releases and many developers seem to run production deployments off the unstable master branch which is not feasible.
-
High risk for bugs in releases due to lack of alpha/beta distribution channels to mature features. New versions currently often come with breaking changes, without properly testing new features through alpha/beta releases to mature them. We acknowledge these shortcoming and are expecting to move to fixed releases later this year.
*There are other DB adapters available such as Postgres, but these are usually less well supported and developed than the adapter for MongoDB.
