Benoit Chesneau
Benoit Chesneau
@willemdj sorry, have been busy these days. I still have to put the code under test. What I did for now is to makes the code of the client more...
@miohtama sure, please send a PR for it :)
i'm not sure to understand the issue... This kind of the point of max_requests to kill the worker if the max is achieved. ``` Command line: --max-requests INT Default: 0...
old issue without any activity. closing it. feel free to open *a new ticket* if needed.
sorry I missed that ticket. I think the intention of this hooks is to ensure this is handled at the right time thois is why it ha s not been...
@tilgovi how do you think we can progress on that topic? Would be good to draft some roadmap for it. In the mean time I would postpone it for the...
postponing to 20.1 let's try to find a solution for it :)
So back on this topic since someone reported on irc he hit the error. The problem I see in monkey patching during preload is that it would make the socket...
The arbiter itself rely on a blocking behaviour for select and signaling. Also it's safer to bind the socket in blocking mode. This is why it's discouraged to patch outside...
I'm not against patching it sooner in the case of preload case. Maybe indeed it won't arm to just patch it, so why not trying it in a branch for...