[V4] Eager-loading performance

[victorLessa]

Lastly I went through some performance problems using Lucid, where it generated a node.js I / O lock preventing other calls from being executed
I came to this conclusion after reading https://nodejs.org/en/docs/guides/dont-block-the-event-loop/
and perform a profilling series in production and using a package to track the IO blocking time
https://github.com/naugtur/blocked-at#readme
(Great package) With this research I was able to notice that the lodash forEach, used by Lucid to carry out the relationships caused a rough block in the IO. Something that is not indicated according to the Node.js documentation

I solved my problem by breaking my query with lucid into smaller charges. Before my queries returned more than 1000 records causing 9 second crashes in the Node.js IO, according to the Naugtur package.
I would like to know if anyone in the community went through this, if so, how did they resolve it?
And I would like to raise a discussion about the possibility of revising Lucid to prevent this from happening on large projects, as is my case, or even adding an alert in the Lucid documentation.

[McSneaky]

Hmm, would be interesting to see, if same thing appears in V5

[victorLessa]

I still don't feel comfortable using adonis version 5 in a big project

[thetutlage]

Hard to comment without looking at the code and the process you used for monitoring the event loop.