Private's questions

I used the gitlab omnibus installtion for my 16.04 Ubuntu Server, since I run into dependency issues when using apt-get install git. So I ran sudo gitlab-rake gitlab:check to check my environment. The check complained that the repo path is not owned by git:git which I don't get... the repo path is owned by git:git and I can create repos. So in general everything works as it should be, but mails are not send out. When running the installation I was asked about my domain and I used my domain without the ending .de. May this cause issues? So in case I want to change anything I have to change it in the gitlab.rb and run a gitlab-ctl reconfigure, correct? I tested my postfix installtion and it can send out mails manually... Port 25 is open. Any Ideas? There is no test mail function in gitlab, or?

I am currently setting up a website including a CMS which has millions of assets. Smaller and bigger pictures mainly. I would like to backup these files on a regular basis. For example weekly. I have an FTP mounted to my machine to which I can copy to. Yesterday I thought about using rsync and simply copy only the new files. But it seems to take a bit long for rsync to crawl through all the files. Basically it takes ages because the FTP storage is shitty. So I ended up with the idea to find only the latest modified (last 24h) files in my assets directory and copy them to the FTP. To minimize the load on the FTP. But I am new to bash scripting, of course google helped me already and I came up with the following parts:

#!/bin/bash

Source="/my/source/folder"
Destination="/my/slow/ftp/"

ls find $Source -mtime -1 -ls
    do
        cp -a $Source $Destination
done    

What am I missing? Can you help me to finish it?

If my idea of backing up the delta is not optimal, feel free to suggest something else.

• I am currently setting up a system which uses a MySQL database.
• I want to create automated backups which run every evening at 10:00PM and also keep the backups for 30 days.

I know there is a way using cronjob and a bashscript which is executed. I already configured a .my.cnf for my SQL user so I won't have to paste the password into the bashscript.

All the tutorials I find do have some extras which I don't need and currently I don't feel to confident to create a script myself.

Do you have an idea or tutorial which is simple and easy?

I created a VM on Azure running Ubuntu 14.04 and installed a LAMP stack. In Azure you have an so called enpoint configuration manager, which basically let you redirect incoming requests from one port to another on the hosted machine. So I deployed my first website on port 81 and made a redirect from 443 to 81 to make it work with the SSL certificate. Now I configured my second website on port 82 which works fine using http, but how can I now bind my SSL to it? Since I have an elastic IP I have to bind my DNS records to the CNAME of the machine which is called for example "mymachine.cloudapp.net:port". I am not so experienced in DNS / SSL / etc. So I think there must be a way which I don't know yet. So far my only idea is to create one machine per website, which for sure is not the best solution. Using google I found plenty of workarounds for IIS but nothing for Apache 2.4...

I installed graphite and grafana which I can both access through web. I was sending some test metrics to the machine using a SSC Serv Utility, which is basically a windows tool where you enter you graphite URL / IP and it starts sending common metrics from performance monitor. CPU / Memory etc.

In my graphite I see all the incoming metrics already but grafana stays empty.

I assume my database settings are not correct, here are my grafana settings:

// Graphite & Elasticsearch example setup

datasources: {
  graphite: {
    type: 'graphite',
    url: "localhost:80",
  },
  elasticsearch: {
    type: 'elasticsearch',
    url: "http://my.elastic.server.com:9200",
    index: 'grafana-dash',
    grafanaDB: true,
  }
},

localhost:80 is the url for my graphite website and I have grafana running on port 81.

What could I be missing? Thx in advance, every help is upvoted.

I have an Ubuntu server 12.04 using a reverse proxy so that tomcat runs on top of apache. And I have a confluence instance on that machine, and a ticket system. So nothing special.

Memory is 8GB that's more than enough for this use.

From time to time, and in the last few days more often my confluence page crashes.

The log (confluence\logs\catalina.2014-05-06) shows this :

SEVERE: Socket accept failed
java.lang.OutOfMemoryError: Java heap space

So how can I avoid this? How can I give that process more memory.

I need to know when my server was rebooted? I only see my last login when I login.

Where is this info stored?

I have an ssl certificate on one server and i am migrating this machine. Can I simply copy paste that certificate? When I have ssl mods installed and enabled?

On the old server I use apache and on the new one we have nginx. At what do I have to pay attention too ?

(old server has ubuntu 10.10 new one has 12.04)

Running Ubuntu 12.04, I setup a private git server and created a group called git some time ago. Now I am following a guide to install gitlab and when it came to adding a user to the git group and create it, I saw that I already had it.

The command is this:

sudo adduser --disabled-login --gecos 'GitLab' git

taken from Gitlab installation Tutorial.

I would like to understand that command correctly. For me I thought I add a user to a group like this:

adduser user group

So what do --disabled-login and --gecos stand for?

I am running Ubuntu 12.04 and created a git server and gave some users access to one repos I created. I have trouble to download the repos using open SSH through command line, no putty nothing.

I am using this command to download the repos called training :

ssh://[email protected]:22/srv/git/training.git and I tried a lot like

ssh://[email protected]:22/srv/git/training.git training   

Using this tutorial : ted felix git server

The error I get is this :

fatal: '/srv/git/training/training' does not appear to be a git repository
fatal: Could not read from remote repository.

My suggestion is, that my repos is not setup correctly.

or I made a mistake with this ssh keys. Any Ideas? I can provide all information you need, just let me know what's missing.

EDIT : It works using the command line (git bash) ... intereseting

Output as requested of folder permissions

XXXXXX@DefaultEnvironment:~$ ls -la /svr/git/training
total 16
drwxr-sr-x 3 root      git 4096 Feb  4 12:55 .
drwxrwsr-x 4 XXXXXX    git 4096 Feb  4 12:40 ..
-rw-r--r-- 1 root      git    9 Feb  4 12:55 catchmeifyoucan
drwxrwsr-x 8 root      git 4096 Feb  4 12:56 .git

right now i am running ruby 1.9 , which i installed using RVM . I needed that to run my redmine 2.4.2 , since i had trouble using 1.8.

Now i need to configure my passenger mod to point to ruby 1.9 but i can't find it.

right now its configured like this :

<IfModule mod_passenger.c>
PassengerRoot /usr
PassengerRuby /usr/bin/ruby
PassengerDefaultUser www-data
</ifModule>

and Since i get the same error as before, i guess it still uses my old 1.8 version. running ubuntu precise 12.04

if you need additional information, please let me know.

I was about to play around on my virtual machine running Ubuntu 12.04. I installed GIT-CORE using apt-get install git-core from repos. First I thought the apt-get command will install in the folder I am in, but I was wrong. So is there a way to directly tell apt-get where to install ?

Or do I have to move the files later? Let me know your best solution for this.

My scenario is like this : I have installed redmine on my ubuntu 12.04 server, and i installed it using apt-get install redmine. So we run the latest stable version 1.3.2 (mysql) , some time after that I found out that there is already 1.4 and even 2.x available to download.

So I wanted to upgrade to 1.4 or any other newer version. But I found out that the newer versions are not available on the official repository.

So from now on I want to check which versions are available in the ubuntu repository . Is there probably even a release schedule?

And I heard that I can download redmine 1.4 from repository if I have ubuntu 13.04 , so I wondered how this works? How to check the repository for any other ubuntu version ?

Many Questions, looking for answers

I am working with Ubuntu servers since some month, so not too long. I never upgraded an ubuntu version. So right now I am running 12.04 and what I need is 13.04. I once hear that I should first upgrade to 12.10 and than to 13.04 (for me that doesn't make too much sense, but I am not sure).

When I run sudo do-release-upgrade -d Ubuntu wants to install 14.04. As far as I know 13.04 is now the stable version. How to achieve the step by step upgrading? Even though it's not the best way, I want to try it out on my virtual machine.

I want to load some modules for proxy configuration. I have the modules installed under /etc/apache2/mods-available/xxx

How do I load them into /etc/apache2/mods-enabled/xxx ?

In other words, in which configuration should this be placed?

LoadModule proxy_module /usr/lib/apache2/modules/mod_proxy.so
LoadModule proxy_http_module /usr/lib/apache2/modules/mod_proxy_http.so 

I tried to locate it with locate imagemagick My result is this :

/etc/bash_completion.d/imagemagick
/usr/lib/mime/packages/imagemagick
/usr/share/doc/imagemagick
/usr/share/doc/imagemagick-common
/usr/share/doc/imagemagick/NEWS.Debian.gz
/usr/share/doc/imagemagick/README.Debian
/usr/share/doc/imagemagick/README.txt.gz
/usr/share/doc/imagemagick/TODO.Debian
/usr/share/doc/imagemagick/changelog.Debian.gz
/usr/share/doc/imagemagick/copyright
/usr/share/doc/imagemagick-common/NEWS.Debian.gz
/usr/share/doc/imagemagick-common/changelog.Debian.gz
/usr/share/doc/imagemagick-common/copyright
/usr/share/menu/imagemagick
/var/cache/apt/archives/imagemagick-common_8%3a6.6.9.7-5ubuntu3.2_all.deb
/var/cache/apt/archives/imagemagick_8%3a6.6.9.7-5ubuntu3.2_amd64.deb
/var/lib/dpkg/info/imagemagick-common.conffiles
/var/lib/dpkg/info/imagemagick-common.list
/var/lib/dpkg/info/imagemagick-common.md5sums
/var/lib/dpkg/info/imagemagick.list
/var/lib/dpkg/info/imagemagick.md5sums
/var/lib/dpkg/info/imagemagick.postinst
/var/lib/dpkg/info/imagemagick.postrm

So which is the path of imagemagick now :( ? I need it to link to it in my mediawiki extension. Any help is upvoted fast !

i gues it's this :

 /usr/lib/mime/packages/imagemagick

but how can I check, or how can I know that ?

I backed up my old database on Ubuntu 10.10 server using MySQL 5.1.49 using mysqldump

Now I want to restore the file on my other machine, using Ubuntu 12.04 server with MySQL version 5.5.32 on it.

I managed to copy the files via rsync but I can't load the DB.

I tried it with:

sudo mysql -u root -p fulldump.sql < fulldump.sql

I get this error:

ERROR 1049 (42000): Unknown database 'fulldump.sql'

BUT the name of the file is fulldump.sql and it's the folder I am currently in of course.

Any ideas?

Simple Question, I want to check if a specific programm = sp is installed on Ubuntu 12.04 server.

I dont want to for sudo apt-get sp and I can't check all folders. Is there a fast and good way?

Right now, I know that I can change permissions for example with:

sudo chmod 550 directory

I wonder how I can check the actual folder permission.

I finally got RVM working and now I would like RVM to install the latest stable ruby version.

In the tutorial I am using they installed ruby 1.9.3 with command :

rvm install 1.9.3

Now I want to know if there is a command to download the latest stable version, or do I have to check latest stable version and download it like that, or else, can I download any old version and run an upgrade?

Normally I would try out everything, but I am working on a new server and I am trying to set it up as clean as I can :)