GitLab 5.4: unable to create user - ruby

(Raised issue https://github.com/gitlabhq/gitlabhq/issues/4659)
I got GitLab 5.4 using Bitnami Stack 5.4.0-0
In this version there's no way to create user without configuring mail server.
(In other words 5.4.0-0 is not ready to use out of box)
Password A temporary password will be generated and sent to user.
User will be forced to change it after first sign in
Where to configure what mail server use for GitLab open source git server?
UPDATE: GitLab 5.4 Installation instruction doesn't mention how to configure connection to mail server, but just how to install Postfix
**Note:** In order to receive mail notifications, make sure to install a
mail server. By default, Debian is shipped with exim4 whereas Ubuntu
does not ship with one. The recommended mail server is postfix and you can install it with:
sudo apt-get install -y postfix
Then select 'Internet Site' and press enter to confirm the hostname.
UPDATE 2: Running sudo apt-get install -y postfix on Bitnami Ubuntu gets
Package postfix is not available, but is referred to by another package. This may mean that package is missing, has been obsoleted, or is only available from another source
E: Package 'postfix' has no installation candidate
UPDATE 3: sudo -u git -H bundle exec rake gitlab:env:info RAILS_ENV=production output
[33mSystem information[0m
System: Debian wheezy/sid
Current User: git
Using RVM: no
Ruby Version: 1.9.3p448
Gem Version: 1.8.24
Bundler Version:1.3.5
Rake Version: 10.0.3
[33mGitLab information[0m
Version: 5.4.0
Revision: 94a814d
Directory: /opt/bitnami/apps/gitlab/htdocs
DB Adapter: mysql2
URL: http://gitlab.funshion.com/gitlab
HTTP Clone URL: http://gitlab.funshion.com/gitlab/some-project.git
SSH Clone URL: git#gitlab.funshion.com:some-project.git
Using LDAP: [32myes[0m
Using Omniauth: no
[33mGitLab Shell[0m
Version: 1.7.0
Repositories: /opt/bitnami/apps/gitlab/repositories/
Hooks: /opt/bitnami/apps/gitlab/gitlab-shell/hooks/
Git: /opt/bitnami/git/bin/git
UPDATE 4: Actually user password can be set if to open the user page again.

We suggest to configure the application to use an external SMTP mail server for sending the emails. If you are using a Cloud image there are limitations for sending emails from the instance itself https://aws-portal.amazon.com/gp/aws/html-forms-controller/contactus/ec2-email-limit-rdns-request In any case, you can send emails from the application if you install "sendmail" in the machine.
There is a quick guide to know how to configure GitLab with GMail or any other email provider easily at http://wiki.bitnami.com/Applications/BitNami_GitLab#How_to_configure_the_email_settings_of_GitLab.3f I hope it helps.

Related

Connect to CentOS desktop (GUI) installed on Google cloud

I've installed Centos7 on the Google Cloud Platform, as you know there is just ssh option to connect to the machine.
Now I need to have access to its desktop(GUI) from my laptop (it's win10).
it's appreciated if anyone can help me.
Thanks in advance
you need to ssh login into your CentOS machine either from ssh button in Gcloud, or using the cloud shell:
gcloud compute ssh test-vws
then setup a password for your account:
sudo passwd `whoami`
After that, you will need to install a client that will allow you to navigate the GUI for your server.
again, on your server command line execute:
sudo yum -y update
sudo yum -y groupinstall 'Server with GUI'
once finished, follow this guige to install and setup Teradici Cloud Access Software (Remote Desktop Agent for CentOS).
Installing Teradici Cloud Access Software
On your virtual workstation, install the Teradici repositories:
sudo yum -y install https://downloads.teradici.com/rhel/teradici-repo-latest.noarch.rpm
Install Teradici Standard Agent for Linux:
sudo yum -y update
sudo yum -y install pcoip-agent-standard
Then you will have to register the Agent and create a firewall rule to allow the connection, just follow the guide, and your desktop connection will look like this:

How to fix docker vulnerabilities (USN-4048-1) in ubuntu 18.08 LTS

I have AWS EC2 ubuntu 18.04 LTS machine. I want to fix the docker vulnerabilities (USN-4048-1).
Installed package : docker.io_17.03.2-0ubuntu7~ppa1
Fixed package : docker.io_18.09.7-0ubuntu1~18.04.3
How can i do that ? What's the steps to fix it ?
The simple way is to follow below step:
If the update-notifier-common package is installed, Ubuntu will alert you about pending updates via the message of the day (motd) upon console or remote login.
After logging in, you can check for and apply new updates with:
When performing an update, first review what apt is going to do, then confirm that you want to apply the updates (this is particularly true when running the development release).
If you would prefer to have updates applied automatically, make sure the unattended-upgrades package is installed, then run 'sudo dpkg-reconfigure unattended-upgrades'. Please note that updates may restart services on your server, so this may not be appropriate for all environments.
$ sudo apt-get update
$ sudo apt-get dist-upgrade
source : https://wiki.ubuntu.com/Security/Upgrades?_ga=2.36307296.1783287754.1567592268-1506938917.1567592268

Permissions problems with openshift client on minishift

I have installed minishift with cask and openshift client with homebrew.
However, as developer user, I am not able to create a project from openshift cli, trigger a build, etc. I always have permissions errors :
Error from server (Forbidden): User "developer" cannot list all project.openshift.io.projectrequests in the cluster
Error from server (Forbidden): User "developer" cannot list authorization.openshift.io.rolebindings in project "my_awesome_project"
Error from server (Forbidden): User "developer" cannot create build.openshift.io.buildconfigs/instantiatebinary in project "my_awesome_project"
However, I am able to create projects from the UI. How can I grant developer user permissions from the command line ?
I am using MacOS High Sierra:
➜ minishift version
minishift v1.15.1+a5c47dd
➜ oc version
oc v3.9.0+191fece
kubernetes v1.9.1+a0ce1bc657
features: Basic-Auth
Server https://192.168.99.100:8443
openshift v1.5.0-rc.0+49a4a7a
kubernetes v1.5.2+43a9be4
You should downgrade your client or upgrade your minishift version to match if possible, as they are not 100% API compatible.
You can use the minishift oc binary in an easy way, so you still preserve you're other client if needed (maybe you need to run multiple versions). To do that do:
localhost# eval $(minishift oc-env)
This will execute:
export PATH="/Users/me/.minishift/cache/oc/v3.7.2/darwin:$PATH"
In this way, you'll always be using the same client version as your minishift instance.
The version of oc that you are using (3.9) is not compatible with the server (1.5 / 3.5). Use a newer version of Openshift or downgrade oc to 3.6.

Could not execute Node.js Cloud9

I am trying to use Cloud9 IDE on a server. I added the SSH key and once I try to SSH into the server the error message
Could not execute node.js on root#xxx.xxx.xxx.xxx
appears.
I have nodejs installed on the server, v0.10.25
You need to install the package "nodejs-legacy".
apt-get install nodejs-legacy
The SSH dialog allows you to set the path to the Node.js binary. That should solve your issue.
I found it is necessary to scroll down and click on the advanced tab. Then I entered /usr/bin/nodejs from my "which nodejs" output in my SSH session. This worked for me even though the documentation states AWS tries to do this by default. That left me in the AWS file work space on the server I ssh'ed into as desired.
nodejs was not installed on my EC2 instance, so I installed using the instructions from the following link, and it work perfectly. Tutorial: Setting Up Node.js on an Amazon EC2 Instance
sudo apt-get install nodejs worked for me

On-Prem Install without using yum command

If datacenter doesn't allow commands like yum, rpm, is there an alternative way to do an on-prem install of OPDK?
They will actually have to install their own local yum repository -- this means building up a rpm repository outside the data center and then brining that into the data center. Then you point yum on the Apigee machine to the internal yum repository.
We have a couple clients who have done this (who probably don't want it advertised so contact me directly for who has done it this way).
You can setup yum client and yum server. Yum client being the machine in the datacenter and yum server which is the machine from which you can pull all required repositories.
Yum will be used to download repositories for open ldap , postgres and qpid.
If for initial testing you are setting up an "sa" ( standalone) installation without analytics you would not require yum. ( If we use analytics all dependencies for analytics(postgres and qpid ) are installed via yum)

Resources