Applications

Send ALT+CTRL+Delete to QEMU virtual machine 1

 in GNU/Linux / Virtualization  tagged alt+ctrl+del / monitor / qemu / sendkey / terminal by

Recently we wanted to start a Windows virtual machine from a physical hard disk using a Fedora w/ GNOME 3 host machine to change the domain password of a user.
To do so, we used QEMU, QEMU is a generic and open source machine emulator and virtualizer.

To perform the password change, we needed to sent the ALT+CTRL+Delete key combination to the virtual machine to access the system screen and then change the user password.
Pressing ALT+CTRL+Delete on the Fedora/GNOME 3 host machine, it popped up a prompt to shut down the host machine instead of sending the key combination to the active window of the VM. Apparently, we could not sent the key combination directly to the VM and had to find a way around it.

Solution:

We pressed ALT+CTRL+2 while the QEMU window was selected/active to switch to the QEMU terminal/monitor.
In the blank screen that appeared, we typed sendkey alt-ctrl-delete and pressed the Enter key.
This action sent to the virtual machine OS the key combination ALT+CTRL+Delete.
Finally, to switch back  to the guest screen we pressed ALT+CTRL+1.

Jenkins: Improve the format of the email 4

 in Applications  tagged jenkins by

Like many people, we use Jenkins to perform several tasks automatically.
Jenkins is an open source automation server, it provides hundreds of plugins to support building, deploying and automating any project.

One of the features of Jenkins is the option to send an email to the user under certain circumstances, e.g. after a build was complete.
The default layout of those emails is not so pretty, so we did some changes to it.
Note: below we present how to make the changes apply to the whole system, if you do not want to do that, you could apply these changes to individual projects only.

Using an administrative account, click on Manage Jenkins option on the left menu and in the new screen click on Configure System on the right column.
These actions will take you to the configuration page of your Jenkins installation (e.g. http://jenkins:8080/configure).

Scroll down to the category Extended E-mail Notification.

  1. From the drop down menu, change the value of Default Content Type to HTML (text/html).
  2. Update the value of Default Subject to [$BUILD_STATUS] - $PROJECT_NAME - Build # $BUILD_NUMBER ($BUILD_ID). We prefer to have the build status first (like a tag).
  3. In the box named Default Content enter the following block of data: 
    $PROJECT_NAME - Build # $BUILD_NUMBER - $BUILD_STATUS.<br/>
<br/>
Check console <a href="$BUILD_URL">output</a> to view full results.<br/>
If you cannot connect to the build server, check the attached logs.<br/>
<br/>
--<br/>
Following is the last 100 lines of the log.<br/>
<br/>
--LOG-BEGIN--<br/>
<pre style='line-height: 22px; display: block; color: #333; font-family: Monaco,Menlo,Consolas,"Courier New",monospace; padding: 10.5px; margin: 0 0 11px; font-size: 13px; word-break: break-all; word-wrap: break-word; white-space: pre-wrap; background-color: #f5f5f5; border: 1px solid #ccc; border: 1px solid rgba(0,0,0,.15); -webkit-border-radius: 4px; -moz-border-radius: 4px; border-radius: 4px;'>
${BUILD_LOG, maxLines=100, escapeHtml=true}
</pre>
--LOG-END--

The email you will receive after a successful execution will be similar to the one below:

Subject: [Successful] - banana - Build # 77 (77)
 Body:

Press the Apply button to save the changes.

You will notice that in the email we mention the following: If you cannot connect to the build server, check the attached logs..
To enable the option to attach the logs, you need to configure your project itself.
Select your project from the main screen and then click on Configure on the left column.

Scroll down to Post-build Actions section.

From the Add post-built action drop down list select Editable Email Notification.

A new block will appear in the page.
Set the value of Attach Build Log drop down to Compress and Attach Build Log and then hit the Apply button.

Back Up Jenkins instance except for workspace and build logs

 in Applications  tagged depth / exclude / find / jenkins / mkdir / rsync by

Our Jenkins setup has a lot of cool features and configuration.
It has ‘project-based security’, it has parametrized projects, multiple source code management blocks per project and fairly extensive tests implemented with several build steps.
Of course, we do not want to lose them, so we make backups often.
The commands we use for the backup are the following.


jenkins_folder="/var/lib/jenkins/";
 backup_folder="$HOME/jenkins/`date +%F`";
 mkdir -p "$backup_folder";
 (cd "$jenkins_folder"/jobs/; find . -mindepth 3 -type d -regex '.*/[0-9]*$' -print) | sed 's|./|jobs/|' | sudo rsync --archive --exclude 'workspace/*' --exclude-from=- "$jenkins_folder" "$backup_folder";

Explanation of commands:

  • In backup_folder="$HOME/jenkins/`date +%F`"; we used the $HOME variable instead of the tilde ~ as this would create a folder in the current directory called ~ instead of creating a new folder called jenkins in the home directory.
  • mkdir -p "$backup_folder"; instructs mkdir to create all parent folders needed to create our destination folder.
  • (cd "$jenkins_folder"/jobs/; find . -mindepth 3 -type d -regex '.*/[0-9]*$' -print) navigates to the directory of jenkins before performing the search, this way the result file names will be relative to the installation location which we need later to pass to rsync.
    Then we search for all folders which their name is numeric and they at least on depth 3. We filter by depth as well to avoid matching folders directly in the jobs folder.
  • sed 's|./|jobs/|' replaces the prefix ./ with jobs/ to match the relative path from where rsync will work from
  • sudo rsync --archive --exclude 'workspace/*' --exclude-from=- "$jenkins_folder" "$backup_folder"; it will copy everything from $jenkins_folder to the folder $backup_folder while excluding the data in workspace and the folders matched from find (the job build folders).
    --exclude-from=- instructs rsync to read from stdin the list of files to exclude.

About restoring a deleted Git branch

 in Applications  tagged filter-branch / git / git-filter-branch / git-reflog / reflog by

Recently, a branch was deleted from the server without it being merged.
Luckily for us, we had a local copy.

We used used the command git reflog to get access to the reference logs of the branch.
The command returned results similar to the below:

271f0084 HEAD@{0}: pull: Merge made by the 'recursive' strategy.
 0e71c0b HEAD@{1}: commit: Minor change: Cracked NSA systems
 c81624b HEAD@{2}: pull: Fast-forward
 ef1f281 HEAD@{3}: commit (amend): Deployed satelite
 1ed1a5a HEAD@{4}: commit (amend): Deployed satelite
 a8682cb HEAD@{5}: commit (amend): Deployed satelite
 e4560c8 HEAD@{6}: commit (amend): Deployed satelite
 1679a90 HEAD@{7}: commit (amend): Deployed satelite
 d27d2c9 HEAD@{8}: commit: Deployed satelite
 aaf8261 HEAD@{9}: checkout: moving from master to crack-nsa-systems
 2500e11 HEAD@{10}: clone: from ssh://[email protected]:7999/secret/bananas.git

From this information we got the hash value in front of the commit which we wanted to use to restore, which was 271f0084.

Then, we checked out that version using

git checkout 271f0084;

When we tried to push the branch back to the server the Git pre commit hooks blocked the operation.
Based on the rsa key used, we could only submit changes that were committed by the owner of that key.
The option to disable temporarily the pre commit hooks was unfortunately out of the question.
So we had to replace all author names and emails with the name of the one holding the key.

To do so we used the following command before pushing to the server once more:

git filter-branch --commit-filter '
   if [ "$GIT_COMMITTER_NAME" = "Doe, John" ];
   then
     GIT_COMMITTER_NAME="Squarepants, Bob";
     GIT_AUTHOR_NAME="Squarepants, Bob";
     GIT_COMMITTER_EMAIL="[email protected]";
     GIT_AUTHOR_EMAIL="[email protected]";
     git commit-tree "$@";
   else
     git commit-tree "$@";
   fi' HEAD

git filter-branch lets you rewrite Git revision history by rewriting the branches mentioned, in our case it was HEAD, applying custom filters on each revision. Those filters can modify each tree (e.g. removing a file or running a perl rewrite on all files) or information about each commit. Otherwise, all information (including original commit times or merge information) will be preserved.

HEAD is a reference to the currently checked out commit. In normal states, it’s actually a symbolic reference to the branch you have checked out.
Looking at the contents of .git/HEAD you’ll see something similar to ref: refs/heads/master.
The branch itself is a reference to the commit at the tip of the branch.