I have just started to understand how node.js event loop and microservices works and I was wondering if microservices can prevent the main thread of node.js application from blocking. What I mean is that we can run synchronous code on a different microservice which can send the response back when done and we can scale only that instance of microservice.
Is my understanding correct or please let me know if I got something wrong?
I think you're mixing up two concepts.
Microservices are relatively small, loosely-coupled services written in whatever language. For example if I work at BigEcommerceCompany I might have a variety of microservices written in a variety of technologies to manage, such as an auth service, cart service, payments service, reviews service, etc., and they might all be in the same language or all be in different languages.
Node's event loop is single threaded, but also has a worker pool that can be used for work without blocking the event loop, and Node can also be clustered with the build in cluster module (or various wrappers) across available CPUs. An example of a blocking function call in Node would be child_process.spawnSync; an example of a non-blocking call would be child_process.spawn. It's common when writing Node code to use a lot of Promises or callbacks to avoid blocking the event loop as much as possible.
The two concepts aren't really related except in that by writing small microservices it may be easier to find, isolate, and fix problems with Node performance.