Re: mod_http2 - h2_session(): connections get's closed on graceful restart

2016-09-01 Thread Stefan Priebe - Profihost AG
Thanks a lot! That was the issue. So it seems it's just easier to trigger with http/2. Greets, Stefan Am 01.09.2016 um 14:45 schrieb Eric Covener: > On Wed, Aug 31, 2016 at 7:37 AM, Stefan Eissing > wrote: >> It really looks as if the CGI process simply died

Re: mod_http2 - h2_session(): connections get's closed on graceful restart

2016-09-01 Thread Stefan Eissing
That would explain things nicely. Just merged it into 2.4.x. > Am 01.09.2016 um 14:45 schrieb Eric Covener : > > On Wed, Aug 31, 2016 at 7:37 AM, Stefan Eissing > wrote: >> It really looks as if the CGI process simply died without sending a (part

Re: mod_http2 - h2_session(): connections get's closed on graceful restart

2016-09-01 Thread Eric Covener
On Wed, Aug 31, 2016 at 7:37 AM, Stefan Eissing wrote: > It really looks as if the CGI process simply died without sending a (part of > a) response. How that is tied to closing stderr or not, I have not the > slightest idea. Can you send me your cgi/perl related

Re: svn commit: r1758672 - in /httpd/httpd/branches/2.2.x: ./ CHANGES include/httpd.h server/protocol.c

2016-09-01 Thread Yann Ylavic
On Thu, Sep 1, 2016 at 10:14 AM, Yann Ylavic wrote: > On Thu, Sep 1, 2016 at 9:57 AM, Yann Ylavic wrote: >> >> I'm not sure the readahead is worth it in 2.2.x though, because >> check_pipeline_flush() does an inconditional flush anyway, and the >>

Re: mod_http2 - h2_session(): connections get's closed on graceful restart

2016-09-01 Thread Stefan Priebe - Profihost AG
Hi, sadly i can't send you that script. It's a big bunch of libs and code and does not run on every server. I'll first try to investigate further. Thanks! Stefan Am 31.08.2016 um 13:37 schrieb Stefan Eissing: > It really looks as if the CGI process simply died without sending a (part of > a)

Re: svn commit: r1758672 - in /httpd/httpd/branches/2.2.x: ./ CHANGES include/httpd.h server/protocol.c

2016-09-01 Thread Yann Ylavic
On Thu, Sep 1, 2016 at 9:57 AM, Yann Ylavic wrote: > > I'm not sure the readahead is worth it in 2.2.x though, because > check_pipeline_flush() does an inconditional flush anyway, and the > issue in 2.4.x was that trailing newline(s) might have caused the next > request to

Re: svn commit: r1758672 - in /httpd/httpd/branches/2.2.x: ./ CHANGES include/httpd.h server/protocol.c

2016-09-01 Thread Yann Ylavic
On Thu, Sep 1, 2016 at 2:51 AM, William A Rowe Jr wrote: > On Wed, Aug 31, 2016 at 2:52 PM, wrote: >> >> Author: ylavic >> Date: Wed Aug 31 19:52:28 2016 >> New Revision: 1758672 >> >> URL: http://svn.apache.org/viewvc?rev=1758672=rev >> Log: >> Merge

Re: Unbounded memory usage in mod_dav + mod_headers/mod_deflate/...

2016-09-01 Thread Christophe JAILLET
Le 29/08/2016 à 18:12, Evgeny Kotkov a écrit : Evgeny Kotkov writes: It might be possible to rework mod_dav_svn, although it's going to take some time. Currently, every top-level handler receives an `ap_filter_t *` and passes it further, and all these places

mod_dav build fails on NetWare with unavailable exports

2016-09-01 Thread NormW
Looking at mod_dav exporting symbols not in use for NetWare, I've (shamelessly) borrowed the same pre-process method used for httpd.imp, which results in the following diff: Index: build/NWGNUmakefile === --- build/NWGNUmakefile