[feature]Let watches Serialize and Parallelize to reduce resource consumption

[huanghe2015]

Version and OS
Any

Is your feature request related to a problem? Please describe.
No

Describe the solution you'd like
As the number of watches grows, the CPU usage by workers(I guess) is also going higher. But in my daily practices, some watches don't need to be known immediately, but I can't determine what interval can I use. Therefore, there is a way as the title mentioned to manage watches and reduce CPU usage: The serialized section and the parallelized section. The parallelized section works as the current version, but the serialized section works as a circular queue, which when a watch is processed, goes to the next.

[dgtlmoon]

This is NOT supported by playwright - already looked into it

If you are not using BrowserSteps, you can do this, set the env var FAST_PUPPETEER_CHROME_FETCHER=1 and it will use pypuppeteer which is a LOT faster and lighter

[huanghe2015]

Thanks for the maintainer's reply, but my advice is just on the watches' level, not the crawler's. So I modified it to watches, not the workers.
I describe my idea in another way: the management of watches is not just by intervals. Some of them can be arranged as a circular queue with just a single thread.