Comment 7 for bug 1251353

Thanks, Mike. Your basic explanation makes a lot of sense of what I have seen so far. I typically take the time to fully understand the source code I am looking at, but I've tried to be upfront about not finding time to do that here, so I also hope my questions haven't seemed too dumb because of that.

Before I saw your reply, I had bumped the value up to 50, and that was still low enough to prevent the instability we had experienced (and certainly far lower than the 600+ plan_level which caused the initial problems). 40 is probably enough, and I am fine with anything up to my test case of 50.

So, I currently have all of this branch's changes in production. Based on that, I am certain it doesn't cause any major problems, but I am less sure we aren't introducing new wrinkles. I know we've talked about it in the past, but do we have any existing tests for this area of the codebase? If not, should we try to cook up a test or two before pushing this?