Difference between revisions of "Gitlab recipes"
From ElphelWiki
(→Solution) |
|||
(7 intermediate revisions by the same user not shown) | |||
Line 10: | Line 10: | ||
==<font color='darkblue'>Nginx behind apache2 misconfiguration</font>== | ==<font color='darkblue'>Nginx behind apache2 misconfiguration</font>== | ||
===Error log=== | ===Error log=== | ||
− | gitlab bind() to 0.0.0.0:80 failed (98: Address already in use) | + | <font size='1em'>gitlab bind() to 0.0.0.0:80 failed (98: Address already in use)</font> |
+ | |||
+ | Also, running '''gitlab-ctl status''' shows that nginx is getting restarted: | ||
+ | <font size='1em'>run: gitaly: (pid 22266) 938s; run: log: (pid 18631) 277569s | ||
+ | run: gitlab-monitor: (pid 22295) 938s; run: log: (pid 18630) 277569s | ||
+ | run: gitlab-workhorse: (pid 22300) 937s; run: log: (pid 18617) 277569s | ||
+ | run: logrotate: (pid 22312) 937s; run: log: (pid 18635) 277569s | ||
+ | run: nginx: (pid 23828) <font color='red'>'''2s'''</font>; run: log: (pid 18629) 277569s | ||
+ | run: node-exporter: (pid 22394) 936s; run: log: (pid 18636) 277569s | ||
+ | run: postgres-exporter: (pid 22401) 936s; run: log: (pid 18649) 277569s | ||
+ | run: postgresql: (pid 22406) 935s; run: log: (pid 18618) 277569s | ||
+ | run: prometheus: (pid 22414) 935s; run: log: (pid 18616) 277569s | ||
+ | run: redis: (pid 22429) 934s; run: log: (pid 18625) 277569s | ||
+ | run: redis-exporter: (pid 22434) 934s; run: log: (pid 18615) 277569s | ||
+ | run: sidekiq: (pid 22440) 933s; run: log: (pid 18626) 277569s | ||
+ | run: unicorn: (pid 22447) 933s; run: log: (pid 18619) 277569s</font> | ||
+ | |||
===Versions=== | ===Versions=== | ||
− | GitLab 10.5.1 | + | <font size='1em'>GitLab 10.5.1 |
GitLab Shell 6.0.3 | GitLab Shell 6.0.3 | ||
GitLab Workhorse v3.6.0 | GitLab Workhorse v3.6.0 | ||
Line 18: | Line 34: | ||
Ruby 2.3.6p384 | Ruby 2.3.6p384 | ||
Rails 4.2.10 | Rails 4.2.10 | ||
− | postgresql 9.6.5 | + | postgresql 9.6.5</font> |
===Problem=== | ===Problem=== | ||
* /etc/gitlab/gitlab.rb: | * /etc/gitlab/gitlab.rb: | ||
− | # nginx['enable'] = true | + | <font size='1em'># nginx['enable'] = true |
# nginx['client_max_body_size'] = '250m' | # nginx['client_max_body_size'] = '250m' | ||
# nginx['redirect_http_to_https'] = false | # nginx['redirect_http_to_https'] = false | ||
− | # nginx['redirect_http_to_https_port'] = 80 | + | # nginx['redirect_http_to_https_port'] = 80</font> |
''redirect_http_to_https'' default option is probably '''true''' or is ignored resulting in a nginx config record: | ''redirect_http_to_https'' default option is probably '''true''' or is ignored resulting in a nginx config record: | ||
* /var/opt/gitlab/nginx/conf/gitlab-http.conf: | * /var/opt/gitlab/nginx/conf/gitlab-http.conf: | ||
− | server { | + | <font size='1em'>server { |
listen *:80; | listen *:80; | ||
... | ... | ||
− | } | + | }</font> |
which is in conflict with Apache | which is in conflict with Apache | ||
===Solution=== | ===Solution=== | ||
− | In ''/etc/gitlab/gitlab.rb'' uncommented and changed '''nginx['redirect_http_to_https_port']''' to another port | + | In ''/etc/gitlab/gitlab.rb'' uncommented and changed '''nginx['redirect_http_to_https_port']''' to another port: |
+ | <font size='1em'>nginx['redirect_http_to_https_port'] = 8033</font> | ||
+ | Then: | ||
+ | <font size='1em'>~# gitlab-ctl reconfigure | ||
+ | ~# gitlab-ctl restart</font> |
Latest revision as of 10:47, 26 February 2018
Contents
X-Frame headers
X-Frame-Options header
- Edit: /opt/gitlab/embedded/service/gitlab-rails/app/controllers/application_controller.rb in def default_headers:
-headers['X-Frame-Options'] = 'DENY' +headers['X-Frame-Options'] = 'ALLOW-FROM https://www.elphel.com/'
- Restart services
gitlab-ctl restart
Nginx behind apache2 misconfiguration
Error log
gitlab bind() to 0.0.0.0:80 failed (98: Address already in use)
Also, running gitlab-ctl status shows that nginx is getting restarted:
run: gitaly: (pid 22266) 938s; run: log: (pid 18631) 277569s run: gitlab-monitor: (pid 22295) 938s; run: log: (pid 18630) 277569s run: gitlab-workhorse: (pid 22300) 937s; run: log: (pid 18617) 277569s run: logrotate: (pid 22312) 937s; run: log: (pid 18635) 277569s run: nginx: (pid 23828) 2s; run: log: (pid 18629) 277569s run: node-exporter: (pid 22394) 936s; run: log: (pid 18636) 277569s run: postgres-exporter: (pid 22401) 936s; run: log: (pid 18649) 277569s run: postgresql: (pid 22406) 935s; run: log: (pid 18618) 277569s run: prometheus: (pid 22414) 935s; run: log: (pid 18616) 277569s run: redis: (pid 22429) 934s; run: log: (pid 18625) 277569s run: redis-exporter: (pid 22434) 934s; run: log: (pid 18615) 277569s run: sidekiq: (pid 22440) 933s; run: log: (pid 18626) 277569s run: unicorn: (pid 22447) 933s; run: log: (pid 18619) 277569s
Versions
GitLab 10.5.1 GitLab Shell 6.0.3 GitLab Workhorse v3.6.0 GitLab API v4 Ruby 2.3.6p384 Rails 4.2.10 postgresql 9.6.5
Problem
- /etc/gitlab/gitlab.rb:
# nginx['enable'] = true # nginx['client_max_body_size'] = '250m' # nginx['redirect_http_to_https'] = false # nginx['redirect_http_to_https_port'] = 80
redirect_http_to_https default option is probably true or is ignored resulting in a nginx config record:
- /var/opt/gitlab/nginx/conf/gitlab-http.conf:
server { listen *:80; ... }
which is in conflict with Apache
Solution
In /etc/gitlab/gitlab.rb uncommented and changed nginx['redirect_http_to_https_port'] to another port:
nginx['redirect_http_to_https_port'] = 8033
Then:
~# gitlab-ctl reconfigure ~# gitlab-ctl restart