A more reasonable way could be to tell the GC to start with a particular (higher than default) heap size. That would avoid a lot of the early GC runs, while still being something that is not outside the bounds that long lasting processes do.
Hi, You are not added the github link into the blog, right?
What I means is, when I read the blog post, there is no way to open the github link in the post, I have to search it in the github manually.