Parse Migration Services

The Parse.com hosting service will be discontinued from January 28,2017. If you are aiming at migrating your app or looking for Parse.com alternatives, you need to start working as soon as possible!

For most apps, the migration process is simple, and will require dedicated development time as well as skilled developer.

We recommend the following schedule for Parse cloud code and Parse hosting:

There are some companies which provide mobile backend as a service. However we recommend you to build your own platform to avoid any issues.

Any cloud service provider should be fine when you are migrating your Parse database from Parse.com cloud code. However, Amazon provides ample amount of ready-to-use services which reduces the cost of development. Keeping all things in Amazon also reduces the cost.

For web servers, one should be using medium capacity servers. These servers should be in auto scaling groups to take benefit of hourly pricing. For databases, we should be using servers with high computing and low storage latency.

Perform the migration in stages.

For example, first, change the notification service from Parse to SNS. Test it and deploy it. In case of database transition, keep the production MongoDB cluster ready before transitioning. Use DB migration tool for transferring data from Parse to your MongoDB cluster. After the DB migration is complete you need to change the DB URIs in your application.

There is unavoidable few minutes of downtime when you complete the DB transition, verify it and change the configured URIs. We can minimise it’s impact by performing the transition during off peak hours.

Use auto-scaling feature of Amazon to increase or decrease the number of servers servicing the APIs. Prefer horizontal scaling instead of vertical scaling wherever possible. Keep the DB and web servers in the same region to reduce the data transfer costs.

Block external accesses to your database servers. Block the ports other than port 80 and 22. Allow SSH access to only restricted IPs.

As frequently as you can. You can also take volume backups. We need to find a balance between the backup frequency and number of days for which the backup is stored. Because the storage cost = number of backups per day * number of days

Yeah. The last question is very important! You have successfully avoided the unnecessary complexities of development until now by using services of Parse.com How do you successfully dodge this bullet? You can either use a service like backend.com or you can use our help to migrate. We will take care of all these decisions for you and deploy your application so that you can continue to steer all your efforts on your application.

Heroku is managed infrastructure and hence, it requires additional costs for management. The AWS and Google cloud are cost effective solutions but one has to manage the software stack, backups, 24x7 availability, downtime, updates, upgrades and security aspects of the hosts created in these environments. Heroku + MongoLab is an alternative to AWS/Google cloud service. Both the approaches have their pros and cons, the ultimate decision is affected by criticality of the application and monthly budget allocated for the parse alternative.