问题
I have GitLab CE setup with my domain (Ubuntu) and the git clone with http not working but ssh works well.
git clone git@gitlab.mydomain.com:anshad/test.git works.
git clone http://gitlab.mydomain.com/anshad/test.git not working.
Cloning into 'test'...
Username for 'http://gitlab.mydomain.com': anshad
Password for 'http://anshad@gitlab.mydomain.com':
fatal: unable to access 'http://gitlab.mydomain.com/anshad/test.git/': The requested URL returned error: 500
sudo -u git -H bundle exec rake gitlab:env:info RAILS_ENV=production says
System information
System: Ubuntu 16.04
Current User: git
Using RVM: no
Ruby Version: 2.3.1p112
Gem Version: 2.5.1
Bundler Version:1.13.6
Rake Version: 10.5.0
Sidekiq Version:4.2.1
GitLab information
Version: 8.14.0-pre
Revision: 358e814
Directory: /home/git/gitlab
DB Adapter: postgresql
URL: http://gitlab.mydomain.com
HTTP Clone URL: http://gitlab.mydomain.com/some-group/some-project.git
SSH Clone URL: git@gitlab.mydomain.com:some-group/some-project.git
Using LDAP: no
Using Omniauth: no
GitLab Shell
Version: 4.0.0
Repository storage paths:
- default: /home/git/repositories/
Hooks: /home/git/gitlab-shell/hooks/
Git: /usr/bin/git
sudo -u git -H bundle exec rake gitlab:check RAILS_ENV=production says
Checking GitLab Shell ...
GitLab Shell version >= 4.0.0 ? ... OK (4.0.0)
Repo base directory exists?
default... yes
Repo storage directories are symlinks?
default... no
Repo paths owned by git:git?
default... yes
Repo paths access is drwxrws---?
default... yes
hooks directories in repos are links: ...
Anshad Vattapoyil / test ... repository is empty
Running /home/git/gitlab-shell/bin/check
Check GitLab API access: OK
Access to /home/git/.ssh/authorized_keys: OK
Send ping to redis server: OK
gitlab-shell self-check successful
Checking GitLab Shell ... Finished
Checking Sidekiq ...
Running? ... yes
Number of Sidekiq processes ... 1
Checking Sidekiq ... Finished
Checking Reply by email ...
Reply by email is disabled in config/gitlab.yml
Checking Reply by email ... Finished
Checking LDAP ...
LDAP is disabled in config/gitlab.yml
Checking LDAP ... Finished
Checking GitLab ...
Git configured with autocrlf=input? ... yes
Database config exists? ... yes
All migrations up? ... yes
Database contains orphaned GroupMembers? ... no
GitLab config exists? ... yes
GitLab config outdated? ... no
Log directory writable? ... yes
Tmp directory writable? ... yes
Uploads directory setup correctly? ... yes
Init script exists? ... yes
Init script up-to-date? ... yes
projects have namespace: ...
Anshad Vattapoyil / test ... yes
Redis version >= 2.8.0? ... yes
Ruby version >= 2.1.0 ? ... yes (2.3.1)
Your git bin path is "/usr/bin/git"
Git version >= 2.7.3 ? ... yes (2.7.4)
Active users: 1
Checking GitLab ... Finished
Apache virtual host configuration for the domain,
<VirtualHost *:80>
ServerAdmin anshad@mydomain.com
ServerName gitlab.mydomain.com
ServerAlias www.gitlab.mydomain.com
ServerSignature Off
ProxyPreserveHost On
AllowEncodedSlashes NoDecode
<Location />
Require all granted
ProxyPassReverse http://localhost:8080
ProxyPassReverse http://gitlab.mydomain.com/
</Location>
RewriteEngine on
RewriteCond %{DOCUMENT_ROOT}/%{REQUEST_FILENAME} !-f [OR]
RewriteCond %{REQUEST_URI} ^/uploads/.*
RewriteRule .* http://127.0.0.1:8080%{REQUEST_URI} [P,QSA,NE]
DocumentRoot /home/git/gitlab/public
LogFormat "%{X-Forwarded-For}i %l %u %t \"%r\" %>s %b" common_forwarded
ErrorLog /var/log/apache2/gitlab_error.log
CustomLog /var/log/apache2/gitlab_forwarded.log common_forwarded
CustomLog /var/log/apache2/gitlab_access.log combined env=!dontlog
CustomLog /var/log/apache2/gitlab.log combined
</VirtualHost>
Config /home/git/gitlab/config/gitlab.yml
production: &base
gitlab:
host: gitlab.mydomain.com
port: 80
https: false
trusted_proxies:
email_from: noreply@mydomain.com
email_display_name: GitLab
email_reply_to: noreply@mydomain.com
email_subject_suffix: ''
default_projects_features:
issues: true
merge_requests: true
wiki: true
snippets: true
builds: true
container_registry: true
incoming_email:
enabled: false
address: "gitlab-incoming+%{key}@gmail.com"
user: "gitlab-incoming@gmail.com"
password: "[REDACTED]"
host: "imap.gmail.com"
port: 993
ssl: true
start_tls: false
mailbox: "inbox"
artifacts:
enabled: true
lfs:
enabled: true
gravatar:
cron_jobs:
stuck_ci_builds_worker:
cron: "0 0 * * *"
expire_build_artifacts_worker:
cron: "50 * * * *"
repository_check_worker:
cron: "20 * * * *"
admin_email_worker:
cron: "0 0 * * 0"
repository_archive_cache_worker:
cron: "0 * * * *"
registry:
gitlab_ci:
ldap:
enabled: false
servers:
main:
label: 'LDAP'
host: '_your_ldap_server'
port: 389
uid: 'sAMAccountName'
method: 'plain' # "tls" or "ssl" or "plain"
bind_dn: '_the_full_dn_of_the_user_you_will_bind_with'
password: '_the_password_of_the_bind_user'
timeout: 10
active_directory: true
allow_username_or_email_login: false
block_auto_created_users: false
base: ''
user_filter: ''
attributes:
username: ['uid', 'userid', 'sAMAccountName']
email: ['mail', 'email', 'userPrincipalName']
name: 'cn'
first_name: 'givenName'
last_name: 'sn'
omniauth:
enabled: false
allow_single_sign_on: ["saml"]
block_auto_created_users: true
auto_link_ldap_user: false
auto_link_saml_user: false
external_providers: []
providers:
shared:
satellites:
path: /home/git/gitlab-satellites/
repositories:
storages: # You must have at least a `default` storage path.
default: /home/git/repositories/
backup:
path: "tmp/backups"
gitlab_shell:
path: /home/git/gitlab-shell/
hooks_path: /home/git/gitlab-shell/hooks/
upload_pack: true
receive_pack: true
git:
bin_path: /usr/bin/git
max_size: 20971520 # 20.megabytes
timeout: 10
extra:
rack_attack:
git_basic_auth:
In /home/git/gitlab/unicorn.rb, I have
worker_processes 4
working_directory "/home/git/gitlab" # available in 0.94.0+
listen "/home/git/gitlab/tmp/sockets/gitlab.socket", :backlog => 1024
listen "127.0.0.1:8080", :tcp_nopush => true
The production.log says only this
Started POST "/ci/api/v1/builds/register.json" for 35.154.26.24 at 2016-11-19 08:16:45 +0000
Started GET "/anshad/test.git/info/refs?service=git-upload-pack" for 137.97.204.60 at 2016-11-19 08:16:47 +0000
Processing by Projects::GitHttpController#info_refs as */*
Parameters: {"service"=>"git-upload-pack", "namespace_id"=>"anshad", "project_id"=>"test.git"}
Filter chain halted as :authenticate_user rendered or redirected
Completed 401 Unauthorized in 23ms (Views: 0.3ms | ActiveRecord: 1.6ms)
Started POST "/ci/api/v1/builds/register.json" for 35.154.26.24 at 2016-11-19 08:16:48 +0000
Started POST "/ci/api/v1/builds/register.json" for 35.154.26.24 at 2016-11-19 08:16:51 +0000
Started POST "/ci/api/v1/builds/register.json" for 35.154.26.24 at 2016-11-19 08:16:54 +0000
Started GET "/anshad/test.git/info/refs?service=git-upload-pack" for 137.97.204.60 at 2016-11-19 08:16:56 +0000
Processing by Projects::GitHttpController#info_refs as */*
Parameters: {"service"=>"git-upload-pack", "namespace_id"=>"anshad", "project_id"=>"test.git"}
Filter chain halted as :authenticate_user rendered or redirected
Completed 401 Unauthorized in 25ms (Views: 0.3ms | ActiveRecord: 2.2ms)
Started GET "/anshad/test.git/info/refs?service=git-upload-pack" for 137.97.204.60 at 2016-11-19 08:16:57 +0000
Processing by Projects::GitHttpController#info_refs as */*
Parameters: {"service"=>"git-upload-pack", "namespace_id"=>"anshad", "project_id"=>"test.git"}
Completed 500 Internal Server Error in 98ms (ActiveRecord: 3.1ms)
JWT::DecodeError (Nil JSON web token):
lib/gitlab/workhorse.rb:120:in `verify_api_request!'
app/controllers/projects/git_http_client_controller.rb:154:in `verify_workhorse_api!'
lib/gitlab/request_profiler/middleware.rb:15:in `call'
lib/gitlab/middleware/go.rb:16:in `call'
Started POST "/ci/api/v1/builds/register.json" for 35.154.26.24 at 2016-11-19 08:16:57 +0000
Issue on GitLab
回答1:
It's important to note that this system referenced in question was built from source code and supported nginx was replaced with Apache (not officially supported by gitlab).
Here is the deal - in the standard nginx config on my system I can see this
upstream gitlab-workhorse {
server unix:/var/opt/gitlab/gitlab-workhorse/socket;
}
proxy_pass http://gitlab-workhorse;
Which means - it's using socket. Not a network port. If I try to see if the workhorse even listening on network - I will see that it's not.
ps -ef|grep -i workhorse
lsof -p pid
Would not show any network ports open by workhorse pid. So perhaps apache config is incorrect? It should be using socket instead of port?
回答2:
First, double-check your gitlab workhorse version and if it is compatible with your current GitLab installation.
Of all the GitLab issues you reference, the comments on 22484 seem the most promising:
In my case, workhorse's logs showed an error accessing
./.gitlab_workhorse_secretAfter some digging, the fix was to add the following to the workhorse startup command line in
/etc/systemd/system/gitlab-workhorse.service:
-secretPath /home/git/gitlab/.gitlab_workhorse_secret
For reference, the full
ExecStartis now:
ExecStart=/home/git/gitlab/bin/daemon_with_pidfile /home/git/gitlab/tmp/pids/gitlab-workhorse.pid \
/home/git/gitlab-workhorse/gitlab-workhorse -listenUmask 0 -listenNetwork unix \
-listenAddr /home/git/gitlab/tmp/sockets/gitlab-workhorse.socket \
-authBackend http://127.0.0.1:8080 -authSocket /home/git/gitlab/tmp/sockets/gitlab.socket \
-documentRoot /home/git/gitlab/public -secretPath /home/git/gitlab/.gitlab_workhorse_secret \
>> /home/git/gitlab/log/gitlab-workhorse.log 2>&1
The other possibility is:
In my case 500 error was caused by bad nginx configuration in
/etc/gitlab/gitlab.rb.
In case where I had something "before" the nginx, like in my case haproxy. I overlooked this fact. It is described in NGiNX settings.
In my case in haproxy sends backend to 8081 where is listening nginx now (originally I setted 8080 - default unicorn service) -
I was not able configure gitlab only with haproxy, without nginx layer.
So in my configuration was important
nginx['listen_port'] = 8081
nginx['listen_https'] = false
Note that both issues are for NGiNX (there is one when Apache2 is used)
There is also a mention about 403 (permission denied) errors:
We were able to resolve the 403 issue by enabling both HTTPS and SSH cloning; we only had SSH cloning enabled which seemed to be causing the problem. This can be changed by going to
https:///admin/application_settingsand double checkingEnabled Git access protocols
Those conclusions are summarized in merge request 6843
But there is more:
Looking at the default files, it looks like there is some sort of confusion with upgrades and what the defaults should be.
With the default configuration file examples (init.dandnginx),gitlab-workhorsewill listen on a Unix socket and not an IP:port.
The Nginx example config file does have some lines for Unix sockets, but the proxy pass goes to an address.I thought setting gitlab up for the first time I needed point my nginx config to the Unicorn port bind as it was the only port I was seeing in netstat get setup when I started the gitlab services.
If you send thegit clonerequest to Unicorn you will get the 500 error.
What I had to do is change gitlab-workhorse to listen to my lookback address and point Nginx there. That cleared up my HTTP 500 error with cloning.
See more with A Brief History of GitLab Workhorse
来源：https://stackoverflow.com/questions/40534682/git-clone-with-http-not-working-in-gitlab