Search
Nginx development mailing list [read only]
Re: Dumping cyclic error_log memory - 4 years ago
Whups, thanks :) A curious point then, why ngx_atomic_fetch_add is used to advance the buffer position, if an atomic operation isn’t necessary since the buffer isn’t shared. Is it a style choice? Sent from my iPhone > On Jan 8, 2020, at 16:50, Datong Sun <datong.sun@konghq.com> wrote: > > > Hello Robert, > > My impression is that the cyclic memory buffer is notby Robert Paprocki - Nginx Development
Re: Dumping cyclic error_log memory - 4 years ago
Hello Robert, My impression is that the cyclic memory buffer is not shared across workers. The memory is allocated using ngx_pnalloc and that is not the shared memory zone. After fork each worker would have their own buffer instead. Which probably explains why this is only suitable for debugging, as you have to know the worker you are trying to debug in order to access the logs this way. Thanksby Datong Sun - Nginx Development
Dumping cyclic error_log memory - 4 years ago
Hello, http://nginx.org/en/docs/debugging_log.html and associated literature instruct to dump the contents of a “memory” error_log target via gdb memory dump. How is it possible to do this for a given nginx worker process, when other worker processes are serving connections simultaneously and generating error logs to the shared memory zone (at least, I’m assuming it’s shared by all workerby Robert Paprocki - Nginx Development
[njs] Fixed [[Prototype]] slot of NativeErrors. - 4 years ago
details: https://hg.nginx.org/njs/rev/84cb3fa5d1cc branches: changeset: 1200:84cb3fa5d1cc user: Dmitry Volyntsev <xeioex@nginx.com> date: Thu Oct 24 16:17:16 2019 +0300 description: Fixed [] slot of NativeErrors. diffstat: src/njs_builtin.c | 86 +++++++++------- src/test/njs_unit_test.c | 227 ++++++---------------------------------------- 2 files changed, 81 inseby Dmitry Volyntsev - Nginx Development
Re: Potential bug in ngx_event.c - 4 years ago
On 2019-09-02 09:24, Maxim Dounin wrote: > My bad, you are right here. > > This isn't a problem though, as accept mutex does not use > sem_wait(), but only uses ngx_shmtx_trylock() - that is, only > atomic operations. > > And for other mutexes we allocate ngx_shmtx_t structures from > shared memory, so it works fine. Got it, very much appreciate the effort looking into tby Rian Hunter - Nginx Development
Re: Potential bug in ngx_event.c - 4 years ago
Hello! On Sun, Sep 01, 2019 at 11:45:20AM -0700, Rian Hunter wrote: > On 2019-08-31 11:28, Maxim Dounin wrote: > > Hello! > > > > On Fri, Aug 30, 2019 at 06:26:57PM -0700, Rian Hunter wrote: > > > >> On 2019-08-30 07:19, Maxim Dounin wrote: > >> > Yes, this is intentional. > >> > > >> > The first agument of the ngx_shmtx_cby Maxim Dounin - Nginx Development
Re: Potential bug in ngx_event.c - 4 years ago
On 2019-08-31 11:28, Maxim Dounin wrote: > Hello! > > On Fri, Aug 30, 2019 at 06:26:57PM -0700, Rian Hunter wrote: > >> On 2019-08-30 07:19, Maxim Dounin wrote: >> > Yes, this is intentional. >> > >> > The first agument of the ngx_shmtx_create() function is a pointer >> > to the ngx_shmtx_t structure, which is not expected to be shared >&gby Rian Hunter - Nginx Development
Re: Potential bug in ngx_event.c - 4 years ago
Hello! On Fri, Aug 30, 2019 at 06:26:57PM -0700, Rian Hunter wrote: > On 2019-08-30 07:19, Maxim Dounin wrote: > > Hello! > > > > On Fri, Aug 23, 2019 at 11:00:41AM -0700, Rian Hunter wrote: > > > >> While browsing the source I noticed something that seemed wrong, even > >> though I haven't observed any buggy behavior or have reproduced this > &by Maxim Dounin - Nginx Development
Re: Potential bug in ngx_event.c - 4 years ago
On 2019-08-30 07:19, Maxim Dounin wrote: > Hello! > > On Fri, Aug 23, 2019 at 11:00:41AM -0700, Rian Hunter wrote: > >> While browsing the source I noticed something that seemed wrong, even >> though I haven't observed any buggy behavior or have reproduced this >> bug >> myself. In ngx_event.c there is a line: >> >> if (ngx_shmtx_create(&by Rian Hunter - Nginx Development
Re: Potential bug in ngx_event.c - 4 years ago
Hello! On Fri, Aug 23, 2019 at 11:00:41AM -0700, Rian Hunter wrote: > While browsing the source I noticed something that seemed wrong, even > though I haven't observed any buggy behavior or have reproduced this bug > myself. In ngx_event.c there is a line: > > if (ngx_shmtx_create(&ngx_accept_mutex, (ngx_shmtx_sh_t *) shared, > cycle->loby Maxim Dounin - Nginx Development
Potential bug in ngx_event.c - 4 years ago
Hello, While browsing the source I noticed something that seemed wrong, even though I haven't observed any buggy behavior or have reproduced this bug myself. In ngx_event.c there is a line: if (ngx_shmtx_create(&ngx_accept_mutex, (ngx_shmtx_sh_t *) shared, cycle->lock_file.data) != NGX_OK) { return NGX_ERROR; } ngx_shmtx_createby Rian Hunter - Nginx Development
[njs] Version 0.3.4. - 4 years ago
details: https://hg.nginx.org/njs/rev/8eadbb3a7c7b branches: changeset: 1127:8eadbb3a7c7b user: Dmitry Volyntsev <xeioex@nginx.com> date: Tue Aug 13 16:13:28 2019 +0300 description: Version 0.3.4. diffstat: CHANGES | 79 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 79 insertions(+), 0 deletions(-) diffs (86 lines): diff -r ab443df0d92by Dmitry Volyntsev - Nginx Development
[njs] Splitting vmcode functionality from njs_vm.c into njs_vmcode.c - 4 years ago
details: https://hg.nginx.org/njs/rev/7fd8df281a37 branches: changeset: 1066:7fd8df281a37 user: Dmitry Volyntsev <xeioex@nginx.com> date: Tue Jul 23 17:53:00 2019 +0300 description: Splitting vmcode functionality from njs_vm.c into njs_vmcode.c No functional changes. diffstat: auto/sources | 1 + njs/njs_core.h | 1 + njs/njs_vm.c | 2435 +----------------by Dmitry Volyntsev - Nginx Development
Re: [PATCH] Mail: nginx_error:cache file has too long header - 5 years ago
Hi Maxim, Sorry for my late reply. We use Nginx as CDN Cache. And it is normal case that the responses are different from different origin servers. For instance: the origin server could be the customer server or some other CDN provider Cache server. The response header could be different. In race condition, I found the inconsistency between disk file and metadata in shared memory. When (c->uby Yongguang Bai - Nginx Development
Re: [PATCH] New directive to configure TLSv1.3 ciphers - 5 years ago
I have a question regarding the ticket. What if different TLSv1.3 ciphers need to be configured in different server blocks? In this case, changing openssl.conf will not help right. Thanks, Ramprasad > On Nov 15, 2018, at 10:55 AM, nginx-devel-request@nginx.org wrote: > > Send nginx-devel mailing list submissions to > nginx-devel@nginx.org > > To subscribe or unsubscribe viaby Ramprasad Tamilselvan - Nginx Development
Re: [PATCH] New directive to configure TLSv1.3 ciphers - 5 years ago
Thanks for the clarification. In this case, I think the patch is not required. Regards, Ramprasad > On Nov 15, 2018, at 9:32 AM, nginx-devel-request@nginx.org wrote: > > Send nginx-devel mailing list submissions to > nginx-devel@nginx.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://mailman.nginx.org/mailman/listinfo/nginx-devel > or, via emaby Ramprasad Tamilselvan - Nginx Development
[nginx] Core: free shared memory on cycle initialization failure. - 5 years ago
details: http://hg.nginx.org/nginx/rev/650574a44505 branches: changeset: 7394:650574a44505 user: Ruslan Ermilov <ru@nginx.com> date: Thu Nov 15 15:28:54 2018 +0300 description: Core: free shared memory on cycle initialization failure. diffstat: src/core/ngx_cycle.c | 64 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 64 insertions(+), 0 deletions(-)by ru@nginx.com - Nginx Development
[nginx] Cache: fixed minimum cache keys zone size limit. - 5 years ago
details: http://hg.nginx.org/nginx/rev/de50fa05fbeb branches: changeset: 7374:de50fa05fbeb user: Maxim Dounin <mdounin@mdounin.ru> date: Wed Oct 31 16:49:39 2018 +0300 description: Cache: fixed minimum cache keys zone size limit. Size of a shared memory zones must be at least two pages - one page for slab allocator internal data, and another page for actual allocations. Usingby Maxim Dounin - Nginx Development
[nginx] Cache: improved keys zone size error reporting. - 5 years ago
details: http://hg.nginx.org/nginx/rev/bddacdaaec9e branches: changeset: 7375:bddacdaaec9e user: Maxim Dounin <mdounin@mdounin.ru> date: Wed Oct 31 16:49:40 2018 +0300 description: Cache: improved keys zone size error reporting. After this change, too small keys zones are explicitly reported as such, much like in the other modules which use shared memory. diffstat: src/httpby Maxim Dounin - Nginx Development
Re: Upstream with zone directive - 5 years ago
> On 17 Sep 2018, at 16:05, Johan Ekenlycka <johan.ekenlycka@tui.se> wrote: > > Hi! > > I am using ngx_lua_upstream module in OpenResty, which is calling ngx_http_get_module_main_conf(r, ngx_http_upstream_module). For some reason, I think the peer data is not correct when the zone directive is used in an upstream. But I am not really sure where in the source the config is paby Sergey Kandaurov - Nginx Development
Re: cache: move open to thread pool - 5 years ago
Hello! On Thu, Aug 09, 2018 at 02:43:19PM -0700, Ka-Hing Cheung via nginx-devel wrote: > On Wed, Aug 8, 2018 at 11:16 AM, Maxim Dounin <mdounin@mdounin.ru> wrote: > > > > We've thought about offloading more operations to thread pools, > > and that's basically why "aio_write" is a separate directive, so > > more directives like "aio_<some_operatiby Maxim Dounin - Nginx Development
Re: cache: move open to thread pool - 5 years ago
Very good feedback, we will look into implementing some of them at least. On Wed, Aug 8, 2018 at 11:16 AM, Maxim Dounin <mdounin@mdounin.ru> wrote: > > We've thought about offloading more operations to thread pools, > and that's basically why "aio_write" is a separate directive, so > more directives like "aio_<some_operation>" can be added in the > fby Ka-Hing Cheung via nginx-devel - Nginx Development
[njs] Setting exception values where appropriate. - 5 years ago
details: http://hg.nginx.org/njs/rev/d2cbea77122c branches: changeset: 582:d2cbea77122c user: Dmitry Volyntsev <xeioex@nginx.com> date: Thu Aug 02 19:39:55 2018 +0300 description: Setting exception values where appropriate. This fixes #42 issue on Github. diffstat: njs/njs_array.c | 21 ++++++++++++++++----- njs/njs_crypto.c | 27 ++++++++++---------------by Dmitry Volyntsev - Nginx Development
Re: Routing to a certain worker - 5 years ago
Hello! On Mon, Jul 23, 2018 at 01:53:21PM -0400, M.P. Ardhanareeswaran wrote: > I understand the desire for worker processes to be symmetric. > > ProxyPass and upstream allow sticky session/session affinity > which does introduce some skew though mighty useful IMHO. > > Do you have any thoughts about use of this? > > I am very new to handler development in NGINX.by Maxim Dounin - Nginx Development
[nginx] Events: moved ngx_recvmsg() to new file src/event/ngx_event_udp.c. - 5 years ago
details: http://hg.nginx.org/nginx/rev/88a624c9b491 branches: changeset: 7285:88a624c9b491 user: Roman Arutyunyan <arut@nginx.com> date: Fri Jun 01 16:55:49 2018 +0300 description: Events: moved ngx_recvmsg() to new file src/event/ngx_event_udp.c. diffstat: auto/sources | 1 + src/event/ngx_event.h | 4 + src/event/ngx_event_accept.c | 357 +----by Roman Arutyunyan - Nginx Development
Re: patch for fixing timouts #189 - 6 years ago
Hello! On Tue, Feb 20, 2018 at 08:51:08PM +0100, Felix Ruess wrote: > quite a while ago someone else already found the problem that the timeouts > use system time rather than a monotonic clock: > https://trac.nginx.org/nginx/ticket/189 > > I updated/fixed the patch [1] there a while ago already and have this > successfully deployed on ~50 devices since then. > > Any chaby Maxim Dounin - Nginx Development
Re: Permanent structure to accumulate data - 6 years ago
Unfortunately, the question is ambiguous. You can put it into the structure, you initialized `mycf = ngx_pcalloc(cf->pool, size)`, where cf is `ngx_conf_t *cf`, if you want to save it per location. Then you can get it in the handler, using `mycf = ngx_http_get_module_loc_conf(r, ngx_http_your_module);`... There are many another places, where you can put it. E. g. just as static variable liby Sergey Brester - Nginx Development
Re: Fix "header too long" error - 6 years ago
> On 16 Jan 2018, at 14:57, Vadim Fedorenko <vadimjunk@gmail.com> wrote: > > # HG changeset patch > # User Vadim Fedorenko <vfedorenko@yandex-team.ru> > # Date 1516103689 -10800 > # Tue Jan 16 14:54:49 2018 +0300 > # Node ID deaa364977488f3390d48306c34dc80961e54e14 > # Parent fbf6a421212b291cbacfcfc503173c0168449165 > Fix “header too long" errorby Sergey Kandaurov - Nginx Development
Re: FW: [PATCH] Using worker specific counters for http_stub_status module - 6 years ago
Hello! On Tue, Jan 09, 2018 at 07:08:41PM +0000, debayang.qdt wrote: > I had this observation while benchmarking nginx 48 workers with > wrk on two separate back to back high speed connected systems > (arm) with several random files being accessed from the client. > As you rightly mentioned this may not have impacted performance > any real time workload in any significant way -by Maxim Dounin - Nginx Development
FW: [PATCH] Using worker specific counters for http_stub_status module - 6 years ago
Hello, I had this observation while benchmarking nginx 48 workers with wrk on two separate back to back high speed connected systems (arm) with several random files being accessed from the client. As you rightly mentioned this may not have impacted performance any real time workload in any significant way - as has been observed during benchmarking. However if it's easy to avoid shared memory contby debayang.qdt - Nginx Development
 |
 |
 |
 |
|