Hi,
On Thu, Jul 11, 2024 at 08:51:54PM +0400, Sergey Kandaurov wrote:
>
> > On 11 Jul 2024, at 13:33, Roman Arutyunyan wrote:
> >
> > Hi,
> >
> > On Mon, May 06, 2024 at 11:14:24AM +0800, ka...@kasei.im wrote:
> >> Hello,
> >>
> >> Thanks for your confirmation and explanations.
> >>
> >> The
> On 11 Jul 2024, at 13:33, Roman Arutyunyan wrote:
>
> Hi,
>
> On Mon, May 06, 2024 at 11:14:24AM +0800, ka...@kasei.im wrote:
>> Hello,
>>
>> Thanks for your confirmation and explanations.
>>
>> The following is a modified patch. In my test it would send GOAWAY(on
>> stream id 0), same as t
Hi,
On Mon, May 06, 2024 at 11:14:24AM +0800, ka...@kasei.im wrote:
> Hello,
>
> Thanks for your confirmation and explanations.
>
> The following is a modified patch. In my test it would send GOAWAY(on
> stream id 0), same as the previous one, but call
> ngx_http_v2_finalize_connection instead o
Hello,
Thanks for your confirmation and explanations.
The following is a modified patch. In my test it would send GOAWAY(on
stream id 0), same as the previous one, but call
ngx_http_v2_finalize_connection instead of set c->close.
# HG changeset patch
# User Kasei Wang
# Date 1714965008 -28800
#
Hi,
On Tue, Apr 23, 2024 at 06:02:08PM +0800, Kasei Wang wrote:
> Hello,
>
> I found that there is a slight probability of HTTP/2 connections not
> properly closing during graceful shutdown, leading to worker processes
> in shutting down state remaining stuck for an extended period. After
> inves
Hello,
I found that there is a slight probability of HTTP/2 connections not
properly closing during graceful shutdown, leading to worker processes
in shutting down state remaining stuck for an extended period. After
investigation, the issue appears to stem from the following:
1. worker processes
