[Support] GitLab-CE


Recommended Posts

5 minutes ago, shaunmccloud said:

Let me easily edit the port mapping after using the template.  Rather than dropping into the CLI.  I did create a pull request to fix the port in the template though.

Do you know unraid have a webgui you can use to easily edit the container? 

Link to comment
21 hours ago, saarg said:

Do you know unraid have a webgui you can use to easily edit the container? 

Yeah, but I'm more used to making changes like that in the CLI or Portainer so I tend to forget about it.  At any rate, the issue is the template is wrong and needs to be fixed.  Pull request is still outstanding.

Link to comment
3 hours ago, shaunmccloud said:

Yeah, but I'm more used to making changes like that in the CLI or Portainer so I tend to forget about it.  At any rate, the issue is the template is wrong and needs to be fixed.  Pull request is still outstanding.

 

I didn't get from your first post that it was the container port that was the problem.

Don't know how active he is, but seeing he hasn't merged the PR, most likely not active.

it's easily fixed when installing through the webui.

 

You don't use CA to install containers?

Link to comment

I'm exposing this to internet via the letsencrypt app also running on my Unraid server.  I have configured a .htaccess file to enable http basic auth so it's not wide open.

I ran into an issue because the name had upper case letters which don't play nice with Docker's DNS.

https://stackoverflow.com/questions/55518144/using-variable-in-nginx-conf

https://github.com/linuxserver/docker-letsencrypt/issues/287

 

Perhaps the name should change to gitlab-ce instead of GitLab-CE

Link to comment
  • 3 months later...

Just updated my docker today to the latest (it was a couple months old) and now it's not starting correctly. I am getting this error in the log:

 

* bash[migrate gitlab-rails database] action run
[execute] rake aborted!
Your current database version is too old to be migrated. You should upgrade to GitLab 11.11.0 before moving to this version. Please see https://docs.gitlab.com/ee/policy/maintenance.html#upgrade-recommendations]

 

Do I need to bash into this docker and manually upgrade the Gitlab core? I had assumed the docker would take care of any upgrades needed. Thanks.

 

Link to comment
9 hours ago, tomahawk1277 said:

Just updated my docker today to the latest (it was a couple months old) and now it's not starting correctly. I am getting this error in the log:

 

* bash[migrate gitlab-rails database] action run
[execute] rake aborted!
Your current database version is too old to be migrated. You should upgrade to GitLab 11.11.0 before moving to this version. Please see https://docs.gitlab.com/ee/policy/maintenance.html#upgrade-recommendations]

 

Do I need to bash into this docker and manually upgrade the Gitlab core? I had assumed the docker would take care of any upgrades needed. Thanks.

 

You can check the dockerhub page for this container and under tags see if they have tagged any versions. If they have, you can just add :the_tag_from_dockerhub at the end of the repository field.

If they don't have versioned tags, you have try the tags one by one until you find one that is old enough.

Link to comment
11 hours ago, saarg said:

You can check the dockerhub page for this container and under tags see if they have tagged any versions. If they have, you can just add :the_tag_from_dockerhub at the end of the repository field.

If they don't have versioned tags, you have try the tags one by one until you find one that is old enough.

That worked, Thanks Saarg! Manually specified a few tagged version numbers from one upgrade to the next until I was able to get the latest. Thanks for the help. 

Link to comment
  • 1 month later...
On 11/26/2018 at 7:06 AM, ElectricBadger said:

I've got this in my nginx config (in proxy-confs/gitlab-ce.subdomain.conf):


server {
  listen *:80;
  server_name  registry.subdomain.duckdns.org;
  server_tokens off;
  return 301 https://$http_host:$request_uri;
  access_log  /config/log/nginx/gitlab_registry_access.log;
  error_log   /config/log/nginx/gitlab_registry_error.log;
}

server {
  listen *:443 ssl;
  server_name  registry.subdomain.duckdns.org;
  server_tokens off;

  include /config/nginx/ssl.conf;
  client_max_body_size 0;
  chunked_transfer_encoding on;

  access_log  /config/log/nginx/gitlab_registry_access.log;
  error_log   /config/log/nginx/gitlab_registry_error.log;

  location / {
    include /config/nginx/proxy.conf;
    proxy_pass http://gitlab:9381;
  }

  location /v2/ {
    include /config/nginx/proxy.conf;
    add_header 'Docker-Distribution-Api-Version' 'registry/2.0' always;
    proxy_pass http://gitlab:9381;
  }
}

and this as the container's extra parameters (minus setup for email, backups and container registry):


--env GITLAB_OMNIBUS_CONFIG="external_url 'https://git.subdomain.duckdns.org';registry_external_url 'https://registry.subdomain.duckdns.org';gitlab_rails['gitlab_ssh_host']='git.subdomain.duckdns.org';nginx['listen_port']=9080;nginx['listen_https']=false;nginx['hsts_max_age']=0;registry_nginx['listen_port']=9381;registry_nginx['listen_https']=false;registry_nginx['enable']=true;registry['enable']=true;"

If you want to use ssh rather than https for cloning/pushing, you'll need to add a port to the container for it: host port 9022 (or whatever you like that isn't being used), container port 22. I then add this in ~/.ssh/config:


Host gitlab
HostName git.subdomain.duckdns.org
Port 9022
User gitlab

although I have to say that, while this works for Mac and Linux/BSD, I have no idea what the equivalent is for Windows!

Thank you for sharing your config. I don't understand, however, what the "registry" means in this case and why you are using port 9381 since the default host port is 9080 (and 9022). 

Link to comment
  • 2 weeks later...
On 9/26/2019 at 5:11 AM, frakman1 said:

Thank you for sharing your config. I don't understand, however, what the "registry" means in this case and why you are using port 9381 since the default host port is 9080 (and 9022). 

The "registry" is because I'm using GitLab's Docker registry rather than pushing to DockerHub (from when my upstream connection was only 0.5Mbps). Port 9381 is to put registry.mydomain on a separate port from git.mydomain so the nginx proxy can redirect accordingly.

 

It's safe to skip both those bits if you don't need your own Docker registry.

Link to comment

Anyone have an issue with there being two instances of port 9022 in the template and it makes the docker not start?

I originally fixed it by removing one, but then Fix Common Problems complained, so I just changed one to 9023.

Just wanted to check on this to see if it's truly the original docker template or if my template was somehow misconfigured by me.

Link to comment
27 minutes ago, teedge77 said:

Anyone have an issue with there being two instances of port 9022 in the template and it makes the docker not start?

I originally fixed it by removing one, but then Fix Common Problems complained, so I just changed one to 9023.

Just wanted to check on this to see if it's truly the original docker template or if my template was somehow misconfigured by me.

Mine just has these three ports: 9080, 9443, 9022 by default.

IMG_1889.JPG

Link to comment

Kind of a noob here, hopefully someone can help me out.

 

Got the docker image from CA and i have most of it working via letsencrypt to my [email protected] and its working well but now i want to try it with SSH to avoid passwords (for school projects) but i simply cannot get it to clone, it just hangs... 

 

changing the inside port to 9023 doesn't seem to change anything and changing inside port to 22 makes the docker simply not open....

 

 

here are some of my configs.... Do i have to open port 22?  

uZ1qfi4.png

 

ZqMahlH.png

Link to comment

Anybody running into issues after updating.  I'm getting this error and I haven't changed anything in a long time.

 

Starting Chef Client, version 14.13.11[0m
resolving cookbooks for run list: ["gitlab"][0m
Synchronizing Cookbooks:[0m
- gitlab (0.0.1)[0m
- package (0.1.0)[0m
- postgresql (0.1.0)[0m
- redis (0.1.0)[0m
- monitoring (0.1.0)[0m
- registry (0.1.0)[0m
- mattermost (0.1.0)[0m
- consul (0.1.0)[0m
- gitaly (0.1.0)[0m
- nginx (0.1.0)[0m
- runit (4.3.0)[0m
- letsencrypt (0.1.0)[0m
- acme (4.0.0)[0m
- crond (0.1.0)[0m
Installing Cookbook Gems:[0m
Compiling Cookbooks...[0m
[0m
================================================================================[0m
[31mRecipe Compile Error in /opt/gitlab/embedded/cookbooks/cache/cookbooks/gitlab/recipes/default.rb[0m
================================================================================[0m

[0mMixlib::Config::UnknownConfigOptionError[0m
----------------------------------------[0m
Reading unsupported config value ‘listen_port’.[0m

Link to comment

My container won't come up after updating, either. I don't see that message — but it's difficult to select things from the log, or even read it, when GitLab outputs so much to the thing. I wish the log viewer had a pause button, or at least didn't keep jumping to the bottom every time there was new output…

 

Update: I don't know what it was doing, but the container came back up by itself after being down for just under 24 hours. Some particularly meaty database migration, perhaps?

 

Edited by ElectricBadger
Added progress update
Link to comment
  • 4 weeks later...

I just cant seem to get the reverse proxy to work :/

I use the linuxserver letsencrypt one, but right now when i access the reverse proxy i get redirected to https://LOKALIP:9080/gitlab/users/sign_in

 

my letsencrypt config is:
 

server {
        listen 80;
        server_name DYNDNS.ddns.net;

if ($scheme = http) {
        return 301 https://$server_name$request_uri;
        }

        listen 443 ssl default_server;
        root /config/www;
        index index.html index.htm index.php;
        server_name DYNDNS.ddns.net;
        ssl_certificate /config/keys/letsencrypt/fullchain.pem;
        ssl_certificate_key /config/keys/letsencrypt/privkey.pem;
        ssl_dhparam /config/nginx/dhparams.pem;
        ssl_ciphers XXXX>
        ssl_prefer_server_ciphers on;
        client_max_body_size 0;
        # This header will prevent search engines from indexing your https pages
        add_header X-Robots-Tag "noindex, nofollow, nosnippet, noarchive";

location /gitlab {
            proxy_pass http://LOKALIP:9080;
            proxy_set_header X-Forwarded-Proto https;
            proxy_set_header X-Forwarded-Ssl on;
        }
}

with a few other, working, configs.

 

In the GitLab Docker config i set:
 

--env GITLAB_OMNIBUS_CONFIG="external_url 'https://DYNDNS.ddns.net/gitlab'; nginx['listen_port'] = 9080; nginx['listen_https'] = false"

 

The GitLab Docker Log gives this Error once in a while:

==> /var/log/gitlab/prometheus/current <==
2019-11-19_13:26:05.41451 level=warn ts=2019-11-19T13:26:05.414Z caller=manager.go:567 component="rule manager" group="GitLab Saturation Ratios" msg="Error on ingesting results from rule evaluation with different value but same timestamp" numDropped=1


==> /var/log/gitlab/prometheus/current <==
2019-11-19_13:26:05.41451 level=warn ts=2019-11-19T13:26:05.414Z caller=manager.go:567 component="rule manager" group="GitLab Saturation Ratios" msg="Error on ingesting results from rule evaluation with different value but same timestamp" numDropped=1

Any Idea how to fix this mess and use Gitlab?

Link to comment
On 11/23/2019 at 9:06 PM, frakman1 said:

I'm using the Gitlab CE docker. For some reason, it thinks that it's URL is http://unraid:9080. This impacts all related services and links. Nothing on my home network knows what unraid is so nothing works. Where do I change it?

edit the Docker, klick on advanced view and change the "extra Parameters" line with your IP

Link to comment
  • 2 weeks later...

I'm able to connect to GitLab and it appears to be running correctly, but an error keeps spamming my logs:

==> /var/log/gitlab/prometheus/current <==
2019-12-04_06:09:35.37530 level=info ts=2019-12-04T06:09:35.375Z caller=main.go:293 msg="no time or size retention was set so using the default time retention" duration=15d
2019-12-04_06:09:35.37547 level=info ts=2019-12-04T06:09:35.375Z caller=main.go:329 msg="Starting Prometheus" version="(version=2.12.0, branch=master, revision=)"
2019-12-04_06:09:35.37555 level=info ts=2019-12-04T06:09:35.375Z caller=main.go:330 build_context="(go=go1.12.13, user=GitLab-Omnibus, date=)"
2019-12-04_06:09:35.37563 level=info ts=2019-12-04T06:09:35.375Z caller=main.go:331 host_details="(Linux 4.18.20-unRAID #1 SMP Fri Nov 23 11:38:16 PST 2018 x86_64 2cc6f067be8c (none))"
2019-12-04_06:09:35.37570 level=info ts=2019-12-04T06:09:35.375Z caller=main.go:332 fd_limits="(soft=40960, hard=40960)"
2019-12-04_06:09:35.37575 level=info ts=2019-12-04T06:09:35.375Z caller=main.go:333 vm_limits="(soft=unlimited, hard=unlimited)"
2019-12-04_06:09:35.37916 panic: runtime error: slice bounds out of range

Has anyone come across this before?

Link to comment

For those of you (like me) who were excited about the announcement that Gitlab now integrates with Sourcegraph, I created a Sourcegraph Unraid template based on this Gitlab template. You can find it in Community Applications under "Sourcegraph". It's my first contribution so please give it a shot and let me know if it works for you or not. 

 

If you want to know more about Sourcegraph, this is a quick screencast

 

 

I had to enable the feature at the Gitlab rails console using this:

gitlab-rails console 

Feature.enable(:sourcegraph)

 

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.