On Wed, Jan 13, 2021 at 10:14 AM Jerome Magnin <[email protected]> wrote: > Apologies for forgetting about adding this in my commit message. Out of > curiosity which use case is currently covered for you ? I did most of my > testing on the number of FD and wasn't able to confirm that only startup > is broken, and reload behaves as expected. Either way I first get > alerted, then warned, and I have a new haproxy process with not so > strict limits.
you are right, there is definitely something weird here. It seems like none of the mentioned scenarios works without your patch, so I wonder how I managed to screw up the situation like this. Because I'm pretty sure it worked at some point, at least I hope ;-) I wondered whether this is not related to http://git.haproxy.org/?p=haproxy.git;a=commit;h=e039f26ba463023282356096ec1cc0d110613e5c where I moved code; but a quick revert seems to show it does not change anything. I hope there is a good explanation outside of "I badly tested things"! I definitely remember testing it in a production scenario, odd! -- William
