Attached is an update that I think sorts out all of the documentation
concerns. I broke this section into paragraphs now that it's getting so
long too.
The only code change is that this now labels the controversial lag here
"average rate limit schedule lag". That way if someone wants to
introduce other measures of rate limit lag, like a more transaction
oriented one, you might call that "average rate limit transaction lag"
and tell the two apart.
The rewritten documentation here tries to communicate that there is a
schedule that acts like it was pre-computed at the start of each client
too. It's not ever adjusted based on what individual transactions do.
I also noted the way this can cause schedule lag for some time after a
slow transaction finishes, since that's the main issue observed so far.
--
Greg Smith 2ndQuadrant US greg@2ndQuadrant.com Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support www.2ndQuadrant.com