Re: No Upstream Proxy Headers

On Mon, Sep 04, 2017 at 08:46:48PM +, Kilian Ries wrote: Hi there, > i'm running a nginx (version: nginx/1.13.1) with two vhosts with exact the > same configuration. The only difference is the upstream section: each vhosts > points to a different upstream server / ip. > "nginx -T" will

Re: Too many redirects

Thank you Francis Daly. It works. On 15 Sep 2017 5:45 am, "Francis Daly" wrote: On Thu, Sep 14, 2017 at 11:54:23PM +0800, tseveendorj wrote: Hi there, > location / { > index index.html; > if ($geoip_country_code != "JP") { return 301

Re: empty user-agent and logs

On Thu, Sep 14, 2017 at 10:57:39PM +0300, Christos Chatzaras wrote: Hi there, > xxx.xxx.xxx.xxx - - [14/Sep/2017:22:47:09 +0300] "GET /index.php HTTP/1.1" > 200 26039 "-" "-" > > There is not difference if there is an empty user-agent or a user-agent with > a dash. > > Any idea it shows the

empty user-agent and logs

curl -A "-" https://hostname/index.php and curl -A "" https://hostname/index.php are logged with: xxx.xxx.xxx.xxx - - [14/Sep/2017:22:47:09 +0300] "GET /index.php HTTP/1.1" 200 26039 "-" "-" There is not difference if there is an

[nginx] HTTP/2: shortened some debug log messages.

details: http://hg.nginx.org/nginx/rev/2bf605c6edf7 branches: changeset: 7108:2bf605c6edf7 user: Maxim Dounin date: Thu Sep 14 19:06:06 2017 +0300 description: HTTP/2: shortened some debug log messages. This ensures slightly more readable debug logs on

[nginx] HTTP/2: added logging of 400 (Bad Request) reasons.

details: http://hg.nginx.org/nginx/rev/d77407baccd1 branches: changeset: 7106:d77407baccd1 user: Maxim Dounin date: Thu Sep 14 19:06:03 2017 +0300 description: HTTP/2: added logging of 400 (Bad Request) reasons. diffstat: src/http/v2/ngx_http_v2.c | 13

[nginx] HTTP/2: fixed debug log about indexed headers.

details: http://hg.nginx.org/nginx/rev/d4b031cf32cf branches: changeset: 7107:d4b031cf32cf user: Maxim Dounin date: Thu Sep 14 19:06:05 2017 +0300 description: HTTP/2: fixed debug log about indexed headers. Previously, "get indexed header" message was logged

Re: nginx cache growing well above max_size threshold

Okay cool, I'll give it a try In our case we do not run http2 on the machines since haproxy runs in front as well (which doesn't support http2) I'll also try enable a bit more verbose logging on one of the machines to see what the logs say Thanks a lot Maxim! Best regards, Lucas Rolff Get

Re: nginx cache growing well above max_size threshold

Hello! On Thu, Sep 14, 2017 at 04:34:09PM +, Lucas Rolff wrote: > I have a minor question, so I have an nginx box using > proxy_cache, it has a key zone of 40 gigabyte (so it can cache > 320 million files), a max_size of 1500 gigabyte for the cache > and the inactive set to 30 days. > >

nginx cache growing well above max_size threshold

Hi guys, I have a minor question, so I have an nginx box using proxy_cache, it has a key zone of 40 gigabyte (so it can cache 320 million files), a max_size of 1500 gigabyte for the cache and the inactive set to 30 days. However we experience that nginx goes well above the defined limit - in

Too many redirects

Hi, I configured http (www, non-www) to https (non-www) and https (www) to https (non-www). It is working fine. But I need to add geoip redirection in location / but I got Too many redirect. This server is behind load balancer. LB is redirecting 80 to 81 and 443 to 80. server {

Пустое значение для arg_name (периодически)

Иногда сталкиваюсь с ситуацией, когда логируемый аргумент POST-запроса прописывается прочерком. Подскажите, пожалуйста, почему такое может быть и на что стоит обратить внимание. -- Sincerely, Dmitry Ansimov freelance system administrator skype: cardinal-gray

Re: [PATCH] Resolver: parse hosts file entries

Hi, On Mon, Aug 21, 2017 at 11:44:42PM -0700, Thibault Charbonnier wrote: > In case this patch interests anyone, here is an updated version of it. > > It is fully backwards compatible; as long as the new 'hostsfile' option > is not set, resolvers will behave as usual. > > I believe it delivers

Re: Http2 enable on all virtul host settings automatically

Hello! On Thu, Sep 14, 2017 at 11:06:30AM +, Junaid Malik wrote: > We recently upgraded Nginx from nginx/1.9.12 - nginx/1.13.2, > details of nginx/1.13.2 supported modules are given below. We > are facing problem of automatic enabling of HTTP2 protocol on > bsa1.example.com as we only

Re: "worker process is shutting down" и ошибки open files

14 сентября 2017 г., 15:05 пользователь ingtar написал: > Некоторые долгие могут. Там конфиг бааальшой %) > reset_timedout_connection on отсутствует. > > Я подцепился стрейсом к зависшему воркеру и там были запросы от websocket. > Их все время держат, гоняя

Re: "worker process is shutting down" и ошибки open files

Некоторые долгие могут. Там конфиг бааальшой %) reset_timedout_connection on отсутствует. Я подцепился стрейсом к зависшему воркеру и там были запросы от websocket. Их все время держат, гоняя ping-pong запросы по ним. Кажется дело в них. worker_shutdown_timeout же по идее должен был в этой

Re: "worker process is shutting down" и ошибки open files

Syntax: keepalive_timeout timeout [header_timeout]; Default: keepalive_timeout 75s; Зачем прописывать дефолтные значения не знаю - там и так 75 секунд у меня прописано 10 секунд, у Вас запросы могут выполняться по 75 секунд? reset_timedout_connection on - это есть в конфиге? 14 сентября 2017

[nginx] Stream: fixed logging UDP upstream timeout.

details: http://hg.nginx.org/nginx/rev/0846dd76a487 branches: changeset: 7105:0846dd76a487 user: Roman Arutyunyan date: Tue Sep 12 13:44:04 2017 +0300 description: Stream: fixed logging UDP upstream timeout. Previously, when the first UDP response packet was not

Re: "worker process is shutting down" и ошибки open files

Да, так и сделали. Про openfiles ошибки ушли, а вот висящие процессы в статусе завершения - остались. Шаги для повторения - поставить nginx, дать нагрузку (5к rps например), сделать reload. Процессы не отпустит, но новые поднимутся. Если убрать нагрузку, то они завершатся через какие-то время

Http2 enable on all virtul host settings automatically

Hello Guys, We recently upgraded Nginx from nginx/1.9.12 - nginx/1.13.2, details of nginx/1.13.2 supported modules are given below. We are facing problem of automatic enabling of HTTP2 protocol on bsa1.example.com as we only enabled http2 on dsa1.example.com. Nginx configurations of both sites

Re: "worker process is shutting down" и ошибки open files

Поставьте больше, например worker_rlimit_nofile 102400; хуже от этого точно не будет 2017-09-13 21:06 GMT+03:00 ingtar : > Да, > > proxy_cache_path /lib/init/rw/proxy_cache levels=1:2 > keys_zone=static:10m max_size=100m; > > Завтра попробую посмотреть дебаг логи