How to detect the bottlenecks of Meteor app infrastructure?

I have just launched an app, and I have once migrated it from Digital Ocean (MUPX) to Meteor Galaxy to make it sure the stack under my app is optimal in terms of performance.

There are still some performance issues, and many of them are probably related with not-so-optimal code. But I have noticed, that there is also a problem with subscriptions, that seem to be much slower than I would have expected. Now I’m trying to find it out, where are the bottlenecks of my setup.

Right now, my Meteor app is hosted on Meteor Galaxy’s developer instance with one container. The database is running free tier on mLab. All the clients are in EU/Scandinavia, but Meteor Galaxy and mLab use US-EAST as their region since it’s the only option in Meteor Galaxy. Could that be a bottleneck? And how about free tier of mLab and developer instance in Meteor Galaxy, are they possible bottlenecks? Does upgrading plans solve things?

I guess that it would be wise to gather some hard data of the current setup using Kadira for example, and then start testing things. But what is the suggested way to get started with performance optimization?

As you mentioned, first get Kadira wired up so you can start to see what’s happening behind the scenes. Then read (multiple times) through Kadira’s Meteor Performance 101 site.

1 Like

Thanks, I’ll definitely check Meteor Performance 101 out!