add instr-fs and decide on its configuration

[trentm]

We should add instr-fs as an option at least for the user to enable.

However, it is debatable if this should be enabled by default. Many getting started guides suggest disabling instr-fs or suggest using requireParentSpan: false. (I have not tested the requireParentSpan: false usage to see if that helps deal with the "many fs spans during startup".)

open-telemetry/opentelemetry-js-contrib#1344 is an old (closed by stalebot) discussion around considering disabling instr-fs by default.

[david-luna]

Upstream has decided to disable it by default in @opentelemetry/auto-instrumenations-node
ref: open-telemetry/opentelemetry-js-contrib#2467

[david-luna]

Many getting started guides suggest disabling instr-fs or suggest using requireParentSpan: false. (I have not tested the requireParentSpan: false usage to see if that helps deal with the "many fs spans during startup".)

I think you meant to set this value to true.since the default is false. It looks like a good compromise since as a dev I see value on knowing the file system operations that happen in a trace. I'll do a draft PR and we can discuss the details

[trentm]

It looks like a good compromise

Now the question is: is it a good enough compromise that EDOT differs from vanilla OTel JS on this.

[trentm]

David and I discussed:

• For GA: let's just follow vanilla OTel JS. The 'fs' use case isn't strong enough to justify subtly differing from vanilla OTel.

[trentm]

I think you meant to set this value to true.

Indeed, you are correct. I had it backwards.