Programming Note

Sunday, December 27, 2015

Happy New Year 2016

New Year 2016 by danil-ko on DeviantArt

Thursday, November 26, 2015

Fedora 23 and Wacom Cintiq as Secondary Display

Fedora 23 recognized most of things automatically for Wacom Cintiq and Krtia works out of box with most settings as default.

However, I could not find how to make Cintiq as a secondary display and able to draw on it directly. There is not UI button for it. Setting the Cintiq as a secondary display only make the Cintiq monitor itself as a secondary, however, the stylus detection still map to the primary screen.

After some Google search and debugging, found following post

http://ragnarb.com/configuring-wacom-cintiq-13hd-with-ubuntu-14-04/

which explains in great detail how to handle it.

However, when I run it, it failed with following error

Cannot find device 'Wacom Cintiq 13HD stylus'.


Cannot find device 'Wacom Cintiq 13HD eraser'.

I then use xsetwacom to find the devices real name

xsetwacom --list devices

It then display as

Wacom Cintiq 13HD Pen stylus id: 14 type: STYLUS 

Wacom Cintiq 13HD Pen eraser id: 15 type: ERASER 

Wacom Cintiq 13HD Pad pad id: 16 type: PAD

So I modified the script as following



#!/bin/bash

#

# Toggles which screen the cintiq is mapped to.

if [ "$(cat ~/.wacom-mapping)" -eq 0 ];

then

    echo 1

    xsetwacom set "Wacom Cintiq 13HD Pen stylus" MapToOutput "HEAD-1"

    xsetwacom set "Wacom Cintiq 13HD Pen eraser" MapToOutput "HEAD-1"

    echo 1 > ~/.wacom-mapping

else

    echo 0

    xsetwacom set "Wacom Cintiq 13HD Pen stylus" MapToOutput "HEAD-0"

    xsetwacom set "Wacom Cintiq 13HD Pen eraser" MapToOutput "HEAD-0"

    echo 0 > ~/.wacom-mapping

fi

Re run the script, and it works this time


[localhost ~]$ wacom-toggle-mapping

1

Sunday, November 8, 2015

Update from Fedora 22 to Fedora 23 with Nvidia Proprietary Driver

I decided to upgrade my own workstation to Fedora 23 from Fedora 23 this weekend. After going through some struggles, I thought I will write down some of my steps so people can reference in case they hit similar obstacles.

Please note that I am not responsible for anything case by the info in the article and the article is provided as is.

Please also note not all info is coming from me. Some of them is found on the referenced article stated in there.

Perform Upgrade

I followed following guide:

https://fedoramagazine.org/upgrading-from-fedora-22-to-fedora-23/

sudo dnf upgrade

sudo dnf install dnf-plugin-system-upgrade

sudo dnf system-upgrade download --releasever=23 --best

The only thing that I did different from the article is that I hit error that mention nvidia drive from rpmfusion cannot be upgraded. So I did following

sudo dnf erase *nvidia

and then redo

sudo dnf system-upgrade download --releasever=23 --best

sudo dnf system-upgrade reboot

After that, I am able to successfully update.

However, after the update is complete, when I tried to install nVidia proprietary driver, but facing following issue:

nVidia Driver is not compatible with Fedora 23 as Fedora 23 includes X.Org 1.18 release candidate version
https://ask.fedoraproject.org/en/question/77564/fedora-23-nvidia-proprietary-driver/

To overcome it, I following article
http://rglinuxtech.com/?p=1598

sudo dnf --allowerasing --releasever=22 downgrade xorg-x11-server-Xorg

sudo cat exclude=xorg-x11* >> /etc/dnf/dnf.conf

Follow the article to install nVidia

http://rpmfusion.org/Howto/nVidia

sudo cat exclude=xorg-x11* >> /etc/dnf/dnf.conf

sudo reboot now

However, after reboot, I found the GDM screen shows up, but no user login prompt. After some research, I found following article

https://ask.fedoraproject.org/en/question/77555/fedora-23-gnome-login-issues/

which seem to indicate that GDM not working with nVidia driver (it seem true, if I remove all nVidia package, the GDM shows up correctly).

As a workaround, I use lightgdm to login
https://ask.fedoraproject.org/en/question/37728/unable-to-login-after-upgrade/

On the login screen, press

ctl + Alt + F2

sudo dnf -y install lightdm system-switch-displaymanager

sudo system-switch-displaymanager lightdm

sudo reboot now

After that, it should have lightdm to prompt login and one should have Fedora 23 + nVidia proprietary driver.

Please note to remove the change on

sudo cat exclude=xorg-x11* >> /etc/dnf/dnf.conf

Once there is an offical X.Org 1.18 release and offical nVidia release to support it. (Also switch back lightdm to gdm if it is fixed.)

I did use nVidia proprietary driver because my nVidia card did not support by open source driver and only have software render, which is too slow to use on Krita.

Saturday, June 20, 2015

OpenShift 3 (Fedora 22 Host + Vagrant with libvirt + OpenShift 3 Docker)

Red Ha recently introduced OpenShift 3, which is different from OpenShift 2 by adopting Docker + Kubernetes. Following doc covers some note I took during experimenting with it as it seem there is still some gaps in online docs that is different from my installation experience (I may just miss something too)

Update (06/27/2015): The latest binary v1.0.0 from OpenShift 3 along with latest CentOS patch (that includes Docker 1.6.2 along with necessary SELinux Policy), its example seem to be working okay. But I found that race condition may happen during a lot of image pull (but maybe it is just because I have a single all in one instance)

The code can be found at here:

https://github.com/openshift/origin

The OS in this docs is Fedora 22 WorkStation version, so some settings will be oriented with Fedora specific settings.

Install Vagrant + libvirt (one may check this doc for more detail http://fedoramagazine.org/running-vagrant-fedora-22/):

sudo dnf install vagrant vagrant-libvirt

Install Vagrant Plugin (details can be found at here: https://github.com/openshift/vagrant-openshift)

vagrant plugin install vagrant-openshift

(Optional: Install Docker for local debug)

sudo dnf install docker

Install nfs-server packages (one may find the detail docs at here: http://www.server-world.info/en/note?os=Fedora_22&p=nfs)

sudo dnf install nfs-utils

Install firewall rule for vagrant + libvirt NFS mount (as vagrant + libvirt use NFS mount for share the data between host and VM, it is a requirement), following commands must be run even if you choose to disable your firewall (I am not sure why, but without following rules, NFS will not happen even if your firewall is off), (detail explanation can be found at here http://fedoramagazine.org/running-vagrant-fedora-22/):

sudo firewall-cmd --permanent --add-service=nfs &&sudo firewall-cmd --permanent --add-service=rpc-bind &&sudo firewall-cmd --permanent --add-service=mountd &&sudo firewall-cmd --reload

Install git (for cloing the repository):

sudo dnf install git

For people who like single line command:

sudo -y dnf install vagrant vagrant-libvirt git docker nfs-utils &&
sudo firewall-cmd --permanent --add-service=nfs &&
sudo firewall-cmd --permanent --add-service=rpc-bind &&
sudo firewall-cmd --permanent --add-service=mountd &&
sudo firewall-cmd --reload && vagrant plugin install vagrant-openshift

Move to a folder where OpenShift source can be stored and clone it down:

git clone https://github.com/openshift/origin.git

Move into the directory and create a CentOS 7 configuration file instead of default Fedora 21 ones (the reason is Fedora 7 seem to be close its end of life, so thought to choose a long support edition)

cd origin
vagrant origin-init --stage inst --os centos7 openshift

Check the configuration:

cat .vagrant-openshift.json | grep centos7

Should Produce following:

"os": "centos7",
"box_name": "centos7_inst",
"box_url": "http://mirror.openshift.com/pub/vagrant/boxes/openshift3/centos7_virtualbox_inst.box"
"box_name": "centos7_base",
"box_url": "http://mirror.openshift.com/pub/vagrant/boxes/openshift3/centos7_libvirt_inst.box"
"ssh_user": "centos"
"ssh_user": "centos"

Proceed to stand out the vagrant box (in the origin folder above):

vagrant up

It should continues to stand out the machine, enter password as needed for the network settings and NFS mount. If it is has errors on NFS mount, please use the above docs and other info to help debug. Once the machine completes success, proceed to ssh into the VM (and type password as necessary):

vagrant ssh

Once ssh into the machine, we can pull the image to ensure they are latest

. /data/src/examples/sample-app/pullimages.sh

Download and start up the necessary binary for OpenShift v3

# Download the binary
curl -L https://github.com/openshift/origin/releases/download/v1.0.0/openshift-origin-v1.0.0-67617dd-linux-amd64.tar.gz | tar xzv

# Change to root user (for below scripts)
sudo su -

# http://fabric8.io/guide/openShiftInstall.html
export OPENSHIFT_MASTER=https://$(hostname -I | cut -d ' ' -f1):8443
echo $OPENSHIFT_MASTER

export PATH=$PATH:$(pwd)

# Create the log directory in advance
mkdir -p /var/lib/openshift/
chmod 755 /var/lib/openshift

# Remove previous generated config if anyrm -rf openshift.local.config

nohup openshift start \
--cors-allowed-origins='.*' \
--master=$OPENSHIFT_MASTER \
--volume-dir=/var/lib/openshift/openshift.local.volumes \
--etcd-dir=/var/lib/openshift/openshift.local.etcd \
> /var/lib/openshift/openshift.log &

tail -f /var/lib/openshift/openshift.log

Please wait until the log is print out listening before proceed to next steps

On may then use the web browser to visit the OpenShift console with VM_ADDRESS:8443 in the host machine, one may found the VM_ADDRESS by using following command in the VM

ifconfig

Should have similar output at eth0 (other ones are for Docker and OpenShift)

eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.121.169 netmask 255.255.255.0 broadcast 192.168.121.255
inet6 fe80::5054:ff:fec6:ba8d prefixlen 64 scopeid 0x20<link>
ether 52:54:00:c6:ba:8d txqueuelen 1000 (Ethernet)
RX packets 184375 bytes 442794303 (422.2 MiB)
RX errors 0 dropped 11 overruns 0 frame 0
TX packets 119825 bytes 15028821 (14.3 MiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

In the Vagrant VM, start following command,

# Link the current config
mkdir -p ~/.kube
ln -s `pwd`/openshift.local.config/master/admin.kubeconfig ~/.kube/config
export CURL_CA_BUNDLE=`pwd`/openshift.local.config/master/ca.crt
sudo chmod a+rwX openshift.local.config/master/admin.kubeconfig
sudo chmod +r openshift.local.config/master/openshift-registry.kubeconfig

# Create local registryoadm registry --create --credentials=openshift.local.config/master/openshift-registry.kubeconfig --config=openshift.local.config/master/admin.kubeconfig
# Check the statusoc describe service docker-registry --config=openshift.local.config/master/admin.kubeconfig

Above should print out docker registry info (please check until Endpoint parameter is no longer <NONE>)

Use following commands to log in as a sample user

oc login --certificate-authority=openshift.local.config/master/ca.crt -u test-admin -p test-admin

Use following commands then to create a javaee sample application with WildFly (please check the original blog: http://blog.arungupta.me/openshift-v3-getting-started-javaee7-wildfly-mysql/)

# Create new project

oc new-project test --display-name="OpenShift 3 WildFly" --description="This is a test sample project to test WildFly on OpenSHift 3"

# Create new app

oc new-app -f https://raw.githubusercontent.com/bparees/javaee7-hol/master/application-template-jeebuild.json

# Trace the build

oc build-logs jee-sample-build-1

The above command should yield something similar to following (may need to wait for few minutes):

I0628 04:40:08.327578 1 sti.go:388] Copying built war files into /wildfly/standalone/deployments for later deployment...
I0628 04:40:08.352066 1 sti.go:388] Copying config files from project...
I0628 04:40:08.353790 1 sti.go:388] ...done
I0628 04:40:16.662515 1 sti.go:96] Using provided push secret for pushing 172.30.6.81:5000/test/jee-sample image
I0628 04:40:16.662562 1 sti.go:99] Pushing 172.30.6.81:5000/test/jee-sample image ...
I0628 04:40:24.410955 1 sti.go:103] Successfully pushed 172.30.6.81:5000/test/jee-sample

One then should be able to use describe service

oc describe services

Output:

NAME LABELS SELECTOR IP(S) PORT(S)
frontend template=application-template-jeebuild name=frontend 172.30.88.128 8080/TCP
mysql template=application-template-jeebuild name=mysql 172.30.250.39 3306/TCP

One may use ssh tunnel to post the website on your local workstation by opening up a new terminal, move into the origin code directory (the ip may be different front above, please check the ip for frontend service and change as necessary)

From local workstation (not the vagrant box), use following

cd origin

vagrant ssh -- -L 9999:172.30.167.74:8080

The IP and Port depends on the output above, 9999 is the local machine available port. One may then open localhost:9999 in local workstation browser and it should open the Java app.

Clean up everything (it will remove everything, please be warned)

oc delete pods,services,projects --all

Shutdown the vagrant boxes by using the command on host workstation:

vagrant halt

Or use the following command on host workstation to destroy the VM (in case something goes wrong):

vagrant destroy --force

Sunday, May 10, 2015

Happy Mother's Day Year 2015

Mother Card 05092015 by danil-ko on DeviantArt

Happy Mother's Day Year 2015.

Completed in Krita.

Following is some of concept art before starting it.Start to try out new ways to color things and correct anatomy problems.

Concept Art 05092015 by danil-ko on DeviantArt

Sunday, December 28, 2014

Happy New Year 2015

Happy New Year 2015 by danil-ko on DeviantArt
Artwork is completed in Krita.

Hope you guys like it and wish you guys happy New Year 2015.

Thanks.

Sincerely,
Danil

Sunday, September 7, 2014

CentOS Docker Image with Tomcat 7

Start to experiment docker. However, most blog posts currently used example with Ubuntu and Tomcat. I am more interest in a combination of Fedora/CentOS/RHEL example with Tomcat 7. So I decide to do my own.

Instead of using a pre-build images, I decide I will experiment with building my own centos image.

Thanks for the contribution on GitHub from https://github.com/blalor/docker-centos-base

I modified it a little.

Following is the step I took (following steps do not assume any OS as base OS may not be centos as some people may want to produce one even without install on the hardware)

The overall summary is following:

Use Vagrant and VirtualBox to setup a minimum centos machine -> Use the minimum virtual machine to build a supermin Docker base image -> Use host to build docker image

Please download vagrant

https://www.vagrantup.com/

Please then download VirtualBox

https://www.virtualbox.org/

Installed both

Execute following

mkdir centos_docker_builder;

cd centos_docker_builder;

vagrant box add centos65-x86_64-20140116 https://github.com/2creatives/vagrant-centos/releases/download/v6.5.3/centos65-x86_64-20140116.box

cat << EOF > Vagrantfile
# -*- mode: ruby -*-
# vi: set ft=ruby :

# Vagrantfile API/syntax version. Don't touch unless you know what you're doing!
VAGRANTFILE_API_VERSION = "2"

Vagrant.configure(VAGRANTFILE_API_VERSION) do |config|
  config.vm.box = "centos65-x86_64-20140116"
#  config.vm.provision "shell", path: "auto_build_setup.sh"
  config.vm.synced_folder ".", "/vagrant", :mount_options => ["dmode=777","fmode=666"]
end
EOF

cat << EOFF > start-tomcat.sh
#!/bin/bash

# From https://github.com/arcus-io/docker-tomcat7

ADMIN_USER=${ADMIN_USER:-admin}
ADMIN_PASSWORD=${ADMIN_PASSWORD:-admin}
MAX_UPLOAD_SIZE=${MAX_UPLOAD_SIZE:- 52428800}

cat << EOF > /opt/apache-tomcat/conf/tomcat-users.xml




EOF

if [ -f "/opt/apache-tomcat/webapps/manager/WEB-INF/web.xml" ];
then
   chmod 664 /opt/apache-tomcat/webapps/manager/WEB-INF/web.xml
   sed -i "s^.*max-file-size.*^\t${MAX_UPLOAD_SIZE}^g" /opt/apache-tomcat/webapps/manager/WEB-INF/web.xml
sed -i "s^.*max-request-size.*^\t${MAX_UPLOAD_SIZE}^g" /opt/apache-tomcat/webapps/manager/WEB-INF/web.xml
fi

/bin/sh -e /opt/apache-tomcat/bin/catalina.sh run
EOFF

wget http://www.carfab.com/apachesoftware/tomcat/tomcat-7/v7.0.55/bin/apache-tomcat-7.0.55.tar.gz

cat << EOF > build_centos.sh
#!/bin/bash

set -e
## Following script is coming from GitHub from https://github.com/blalor/docker-centos-base
## Thanks for the code

## requires running as root because filesystem package won't install otherwise,
## giving a cryptic error about /proc, cpio, and utime.  As a result, /tmp
## doesn't exist.
[ $( id -u ) -eq 0 ] || { echo "must be root"; exit 1; }

tmpdir=$( mktemp -d )
trap "echo removing ${tmpdir}; rm -rf ${tmpdir}" EXIT

febootstrap \
    -u http://mirrors.mit.edu/centos/6.5/updates/x86_64/ \
    -i centos-release \
    -i yum \
    -i iputils \
    -i tar \
    -i which \
    -i http://mirror.pnl.gov/epel/6/i386/epel-release-6-8.noarch.rpm \
    centos65 \
    ${tmpdir} \
    http://mirrors.mit.edu/centos/6.5/os/x86_64/

febootstrap-run ${tmpdir} -- sh -c 'echo "NETWORKING=yes" > /etc/sysconfig/network'

## set timezone of container to UTC
febootstrap-run ${tmpdir} -- ln -f /usr/share/zoneinfo/Etc/UTC /etc/localtime

febootstrap-run ${tmpdir} -- yum clean all

## xz gives the smallest size by far, compared to bzip2 and gzip, by like 50%!
febootstrap-run ${tmpdir} -- tar -cf - . | xz > centos65.tar.xz
EOF

chmod a+x build_centos.sh

cat << EOF > Dockerfile
From scratch

MAINTAINER Danil Ko

ADD centos6.5.tar.xz /

# Need to update additional selinux packages due to https://bugzilla.redhat.com/show_bug.cgi?id=1098120
RUN yum -y install yum install java-1.7.0-openjdk-devel wget http://mirror.centos.org/centos/6.5/centosplus/x86_64/Packages/libselinux-2.0.94-5.3.0.1.el6.centos.plus.x86_64.rpm http://mirror.centos.org/centos/6.5/centosplus/x86_64/Packages/libselinux-utils-2.0.94-5.3.0.1.el6.centos.plus.x86_64.rpm

# Use copy to preserve the the tar file without untar
COPY apache-tomcat-7.0.55.tar.gz /tmp/

RUN cd /opt; tar -xzf /tmp/apache-tomcat-7.0.55.tar.gz; mv /opt/apache-tomcat* /opt/apache-tomcat; rm /tmp/apache-tomcat-7.0.55.tar.gz;

ADD start-tomcat.sh /usr/local/bin/start-tomcat.sh

EXPOSE 8080

CMD ["/bin/sh", "-e", "/usr/local/bin/start-tomcat.sh"]

EOF

vagrant up

vagrant ssh "cd /vagrant; . build_centos.sh"

vagrant destroy

# Move the  centos_docker_builder folder to a docker host machine, in this example, it is the same machine
# Also download the tomcat file and name it as apache-tomcat-7.0.55.tar.gz

docker build -t centosimage .

# Run as a background daemon process and with the container name webtier
docker run -d --name webtier centosimage

# On a separate terminal, run (consider one only have the webtier running, otherwise, one will use docker ps -a to find the container id and use docker inspect )

docker inspect -f '{{ .NetworkSettings.IPAddress }}' webtier | grep IP
        "IPAddress": "172.17.0.60",
        "IPPrefixLen": 16,

# On a web browser, do

172.17.0.60:8080

The tomcat webpage will show up

To clean up the container, one may run

docker stop webtier; docker rm webtier;

To clean up the image, after the above command one may run

docker rmi centosimage;

Sunday, July 13, 2014

OpenShift Git History CLean Up

Recently, I found that git history start to fill up the allowed disk space in the gear as I was always deployed the build war files into the gear through git. And git store the entire binary delta every time when there is a new push.

So I start to look around for an ultimate solutions for the problem.

The first solution I encounter is from OpenShift online answer that describe how to clean up remote repo within the gear

https://www.openshift.com/forums/openshift/how-to-erase-all-history-from-a-git-repository-on-openshift-and-start-over-with

This solution works rather well. It removes the entire history. So I used for quite some time.

#!/bin/bash



# Clean up gear info

ssh $gear_ssh_url "cd git; rm -rf *.git; cd *.git; git init --bare; exit"

# Create bare repo to overwrite history (or do a new git clone, it is just I found it is faster just do a new repo and point remote repo)

git init

current_date=`date`

git commit -m "Automatic Push as of $current_date due to code change"

git add origin $gear_git_url

git push origin master -u --force

However, recently, I start to find that I should at least keep one history of backup, so I can rollback more easily (I found that binary-deployment seem to be harder to management then the git, maybe I did not understand the full feature yet). So I start to look for a git solution to clean up history.

Following is the other solution I found

http://stackoverflow.com/questions/11929766/how-to-delete-all-git-commits-except-the-last-five

It works very well. As it do remove the commit history and there reduce the size but also leave at least one history commit that I can revert back to.

What I found is that there is one thing I do not recognize in the solution, it will only impact local repository until there is a new commit to push the change along with these new repo data clean up to effect local repo. So without a new commit, if somebody clone it again, all history will still be there as the change is not in repo yet.

So I did some adjustment

#!/bin/bash




# Reference from the article http://stackoverflow.com/questions/11929766/how-to-delete-all-git-commits-except-the-last-five

current_branch="$(git branch --no-color | cut -c3-)" ;

current_head_commit="$(git rev-parse $current_branch)" ;

echo "Current branch: $current_branch $current_head_commit" ;

# A B C D (D is the current head commit), B is new_history_begin_commit

new_history_begin_commit="$(git rev-parse $current_branch~1)" ;

echo "Recreating $current_branch branch with initial commit $new_history_begin_commit ..." ;

git checkout --orphan new_start $new_history_begin_commit ;

git commit -C $new_history_begin_commit ;

git rebase --onto new_start $new_history_begin_commit $current_branch;

git branch -d new_start ;



git reflog expire --expire=now --all;

git gc --prune=now;



# Still require a push for remote to take effect, otherwise the push will not go through as there is no change

if [ -f .invoke_update ];

then

      rm -rf .invoke_update;

else

      touch .invoke_update;

fi


git add -A .;

current_date=`date`;

git commit -m "Force clean up history $current_date";

git push origin master --force;

It first did as in post to clean up the local repo, make a dummy commit, push the change to remote.

Thanks.

Sincerely,
Danil

Sunday, May 18, 2014

Happy Mother's Day 2014

Happy Mother Day 2014 by danil-ko on deviantART

After a week of delay, I finally ended up finish it.

Completed in MyPaint, Krita.

Thanks.

Sincerely,
Danil

Monday, February 17, 2014

Setup Eclipse Binary for Eclipse Official Website in Fedora 20

Following is a sample note to setup eclipse binary from Official Eclipse Website instead of using yum.

I chose to go down this path since yum is several revisions behind the official version.

# Install jdk development if not already, to prevent some strange behavior when compiling code
sudo yum -y install java-1.7.0-openjdk-devel.x86_64

# Download the eclipse from official site. (http://www.eclipse.org/)
wget http://www.eclipse.org/downloads/download.php?file=/technology/epp/downloads/release/kepler/SR1/eclipse-jee-kepler-SR1-linux-gtk-x86_64.tar.gz eclipse.tar.gz

# Unpackage it
mkdir -p eclipse
cd eclipse
tar -xvf ../eclipse.tar.gz

# Create eclipse.desktop file
cat > ~/eclipse.desktop <<"EOF"
[Desktop Entry]
Encoding=UTF-8
Name=Eclipse IDE
Exec=/opt/eclipse/eclipse
Icon=/opt/eclipse/icon.xpm
Type=Application
Categories=Development;
EOF

# Make it all permission
chmod a+wrx eclipse.desktop

cd ..

# Copy the binary to opt
sudo mv eclipse /opt/eclipse

# Make the desktop symbolic link
sudo ln /opt/eclipse/eclipse.desktop /usr/share/applications/eclipse.desktop

One should be able to see it in Search Menu

Thanks.

Sincerely,
Danil

Sunday, February 2, 2014

Happy New Year 2014

New Year 2014 by danil-ko on deviantART

Artwork is completed in Krita with GIMP to add texts and some enhancements.

Hope you guys like it and wish you guys happy New Year 2014.

Sincerely,
Danil

Sunday, January 26, 2014

OpenShift Origin 3.0 with JBossews CentOS Installation

One can download the CentOS Minimal Image at
http://wiki.centos.org/Download

One of link is
http://mirror.trouble-free.net/centos/6.5/isos/x86_64/

Install the VM Image by using VirtualBox
https://www.virtualbox.org/

It is recommended to use
4GB Memory

40GB Dynamic Disk Space

Following needed to be set
Bridge Adapter

After the installation complete, run following bash script

(Host at github https://github.com/danilko/openshift-origin-3-jbossews-cartridge-installation)

#!/bin/bash

BROKER_OPENSHIFT_HOSTNAME=`hostname`
APP_OPENSHIFT_HOSTNAME=`echo ${BROKER_OPENSHIFT_HOSTNAME} | cut -d '.' -f2-`

# Make eth0 adapater enable during system start up
sed -i 's^ONBOOT=no^ONBOOT=yes^g' /etc/sysconfig/network-scripts/ifcfg-eth0

# Restart network service
service network restart

# Make current hostname to be resolvable
echo  -e "\n127.0.0.1 broker.platform.local" >> /etc/hosts

yum -y install java-1.7.0-openjdk-devel wget unzip

cd /tmp
wget http://dl.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm
rpm -Uvh epel-release-6*.rpm;
cat > /etc/yum.repos.d/openshift-origin-deps.repo <<"EOF"
[openshift-origin-deps]
name=OpenShift Origin Dependencies - EL6
baseurl=http://mirror.openshift.com/pub/origin-server/release/3/rhel-6/dependencies/$basearch/
gpgcheck=0
EOF

yum install -y ruby193-ruby unzip curl bind httpd-tools puppet augeas

# Install default config file
cat > ~/.openshift/oo-install-cfg.yml <<"EOF"
---
Version: 0.0.1
Description: This is the configuration file for the OpenShift Installer.
Deployment:
  DNS:
    component_domain: broker.platform.local
    register_components: Y
    app_domain: platform.local
  Hosts:
  - ip_addr: 127.0.0.1
    named_ip_addr: 127.0.0.1
    user: root
    host: broker.platform.local
    state: new
    ip_interface: eth0
    roles:
    - msgserver
    - dbserver
    - broker
    - node
    ssh_host: localhost
Vendor: OpenShift Origin Community
Subscription:
  type: yum
  jenkins_repo_base: http://pkg.jenkins-ci.org/redhat
  repos_base: https://mirror.openshift.com/pub/origin-server/release/3/rhel-6
Name: OpenShift Installer Configuration
EOF

sh <(curl -s https://install.openshift.com/) -w origin_deploy

# Update the DNS of current host
nsupdate -k /var/named/Kplatform.local*.key
server ${BROKER_OPENSHIFT_HOSTNAME}
update delete ${BROKER_OPENSHIFT_HOSTNAME} A
update add  180 A 127.0.0.1
send
quit

######################################
## SSL Cert Generation
######################################

mkdir -p /tmp/pki
cd /tmp/pki

cat >  platform.crt.config <<"EOF"
 RANDFILE               = $ENV::HOME/.rnd

 [ req ]
 default_bits           = 1024
 default_keyfile        = keyfile.pem
 distinguished_name     = req_distinguished_name
 attributes             = req_attributes
 prompt                 = no
 output_password        = changeme

 [ req_distinguished_name ]
 C                      = US
 ST                     = CA
 L                      = SAN JOSE
 O                      = PLATFORM
 OU                     = PLATFORM
 CN                     = *.platform.local
 emailAddress           = changeme

 [ req_attributes ]
 challengePassword              = changeme
EOF

# Generate a new key that lasts 365 days
openssl req -batch -x509 -nodes -days 365 -newkey rsa:2048 -keyout platform.key -out platform.crt -config platform.crt.config

# Backup old certificate and key
\cp -rf  /etc/pki/tls/certs/localhost.crt /etc/pki/tls/certs/localhost.crt.bak
\cp -rf  /etc/pki/tls/private/localhost.key /etc/pki/tls/private/localhost.key.bak

# Copy certificate and key
\cp -rf platform.crt /etc/pki/tls/certs/localhost.crt
\cp -rf platform.key /etc/pki/tls/private/localhost.key

# Setup so the serveralias also contain *platform.local
sed -i 's^ServerAlias localhost^ServerAlias localhost *.platform.local^g' /etc/httpd/conf.d/000001_openshift_origin_node.conf
sed -i 's^ServerAlias localhost^ServerAlias localhost *.platform.local^g' /etc/httpd/conf.d/000002_openshift_origin_broker_proxy.conf

# Get certificate through opnenssl
# Do not need since the cert is auto generated
#echo -n | openssl s_client -connect broker.platform.local:443 | sed -ne '/-BEGIN CERTIFICATE-/,/-END CERTIFICATE-/p' > platform.crt

# Import to java keystore to prevent ssl problem
keytool -import -trustcacerts -keystore /usr/lib/jvm/java-1.7.0-openjdk-1.7.*.x86_64/jre/lib/security/cacerts -storepass changeit -noprompt -alias platform_cert -file /tmp/pki/platform.crt

cd  /tmp
rm -rf /tmp/pki

# Restart httpd
service httpd restart

# Ensure cgconfig and cgred is on to make sure gear config
/sbin/chkconfig cgconfig on
/sbin/chkconfig cgred on

reboot now

After the reboot, run the following bash script

#!/bin/bash

cd /tmp

TOMCAT7_VER=7.0.50
TOMCAT6_VER=6.0.37
MAVEN_VER=3.1.1

# Download binary
wget http://supergsego.com/apache/tomcat/tomcat-7/v${TOMCAT7_VER}/bin/apache-tomcat-${TOMCAT7_VER}.tar.gz
wget http://apache.mirrors.hoobly.com/tomcat/tomcat-6/v${TOMCAT6_VER}/bin/apache-tomcat-${TOMCAT6_VER}.tar.gz 
wget http://apache.petsads.us/maven/maven-3/3.1.1/binaries/apache-maven-${MAVEN_VER}-bin.tar.gz 
wget https://github.com/openshift/origin-server/archive/openshift-origin-release-3.zip 

# Untar/Unzip binaries
tar -xvf apache-tomcat-${TOMCAT7_VER}.tar.gz
tar -xvf apache-tomcat-${TOMCAT6_VER}.tar.gz
tar -xvf apache-maven-${MAVEN_VER}-bin.tar.gz
unzip openshift-origin-release-3.zip

######################################
## Install Maven
######################################
\cp -rf apache-maven-${MAVEN_VER} /etc/alternatives/maven

# Setup maven path system wide
echo -e 'export M2_HOME=/etc/alternatives/maven\nexport PATH=${M2_HOME}/bin:${PATH}'  > /etc/profile.d/maven.sh

source /etc/profile.d/maven.sh

mvn -v

######################################
## Install JBossews
######################################
# Copy jbossews/tomcat to correct location
\cp -rf apache-tomcat-${TOMCAT6_VER} /etc/alternatives/jbossews-1.0
\cp -rf apache-tomcat-${TOMCAT7_VER} /etc/alternatives/jbossews-2.0

# Copy openshift cartridges
\cp -rf origin-server-openshift-origin-release-3/cartridges/openshift-origin-cartridge-jbossews /usr/libexec/openshift/cartridges/jbossews

chmod a+x /usr/libexec/openshift/cartridges/jbossews/bin/*

# Install the cartridge
oo-admin-cartridge -a install -s /usr/libexec/openshift/cartridges/jbossews

# Clear broker cache
oo-admin-broker-cache -c

# Clean up folder
rm -rf origin-server-openshift-origin-release-3
rm -rf apache-maven-${MAVEN_VER}
rm -rf  apache-tomcat-${TOMCAT7_VER}
rm -rf  apache-tomcat-${TOMCAT6_VER}

######################################
## Change Default User Gear Size all to medium
######################################

# Resource limit file
RESOURCE_LIMIT_CONFIG_FILE=/etc/openshift/resource_limits.conf
sed -i 's^node_profile=small^node_profile=medium^g' ${RESOURCE_LIMIT_CONFIG_FILE}
sed -i 's^quota_blocks=1048576^quota_blocks=2097152^g' ${RESOURCE_LIMIT_CONFIG_FILE}
sed -i 's^quota_files=80000^quota_files=999999^g' ${RESOURCE_LIMIT_CONFIG_FILE}
sed -i 's^memory_limit_in_bytes=536870912       # 512MB^memory_limit_in_bytes=1073741824       # 1024MB^g' ${RESOURCE_LIMIT_CONFIG_FILE}
sed -i 's^memory_memsw_limit_in_bytes=641728512 # 512M + 100M (100M swap)^memory_memsw_limit_in_bytes=1178599424 # 1024M + 100M (100M swap)^g' ${RESOURCE_LIMIT_CONFIG_FILE}

# Set the file connection limit
RESOURCE_LIMIT_CONFIG_FILE=/etc/security/limits.conf
sed -i 's^#End of file^^g' ${RESOURCE_LIMIT_CONFIG_FILE}

echo -e "\n* soft nofile 32000\n* hard nofile 32000" >> ${RESOURCE_LIMIT_CONFIG_FILE}
echo -e "\n#End of file" >> ${RESOURCE_LIMIT_CONFIG_FILE}

# Reboot node services
service mcollective restart
oo-cgroup-enable --with-all-containers
oo-pam-enable --with-all-containers
oo-admin-ctl-tc restart

sed -i 's^"small"^"medium"^g' /etc/openshift/broker.conf

# Change demo user gear size
oo-admin-ctl-user --removegearsize small -l demo
oo-admin-ctl-user --addgearsize medium -l demo

# Clear broker cache
oo-admin-broker-cache -c

######################################
## Setup rhc tool
######################################

# Setup rhc tool
gem install rhc
echo yes | rhc setup --server=broker.platform.local -l demo -p changeme -k –no-create-token

By the end, one should have a working OpenShift configuration with JBossews Cartridges enable

Thanks.

Sincerely,
Danil

Sunday, February 10, 2013

Jersey + Multipart + Maven

I thought I should write about a problem that takes a long time for me to solve in case I forget about it.

I looked through tutorials and other articles to learn about Jersey Multipart. However, I ran into issues with missing dependency by following these tutorials. I finally solved it by setting up my maven POM file in the following way:

<dependencies>
<dependency>
<groupId>junit</groupId>
<artifactId>junit</artifactId>
<version>3.8.1</version>
<scope>test</scope>
</dependency>
<dependency>
<groupId>com.sun.jersey</groupId>
<artifactId>jersey-server</artifactId>
<version>1.9</version>
</dependency>
<dependency>
<groupId>com.sun.jersey.contribs</groupId>
<artifactId>jersey-multipart</artifactId>
<version>1.9</version>
</dependency>
<dependency>
<groupId>com.sun.jersey</groupId>
<artifactId>jersey-json</artifactId>
<version>1.9</version>
</dependency>
<dependency>
<groupId>org.jvnet.mimepull</groupId>
<artifactId>mimepull</artifactId>
<version>1.9</version>
</dependency>
</dependencies>

In short, problems seem to come from version differences across jersey libraries as well as mimepull libraries. All libraries should be the same version. So even if other Jersey libraries already had 1.14 version, it should be safer to use 1.9 version since it was the latest version for mimepull.

The missing dependency error messages were not too informative at this case.

Thanks to authors of original articles and their comments who contribute to the solution.

Hope it helps others.

中文翻譯:

最近花了不少時間解決了一個問題, 想說先把解決方案寫下來以免忘記.

之前幾日在網路上尋找關於Jersey Multipart的教學. 在讀完許多教學後自己寫了一個簡單的範例.但是在Deploy時卻遇到missing dependency的問題. 最後終於在把Maven的POM檔做出以下修改後解決了問題:

<dependencies>
<dependency>
<groupId>junit</groupId>
<artifactId>junit</artifactId>
<version>3.8.1</version>
<scope>test</scope>
</dependency>
<dependency>
<groupId>com.sun.jersey</groupId>
<artifactId>jersey-server</artifactId>
<version>1.9</version>
</dependency>
<dependency>
<groupId>com.sun.jersey.contribs</groupId>
<artifactId>jersey-multipart</artifactId>
<version>1.9</version>
</dependency>
<dependency>
<groupId>com.sun.jersey</groupId>
<artifactId>jersey-json</artifactId>
<version>1.9</version>
</dependency>
<dependency>
<groupId>org.jvnet.mimepull</groupId>
<artifactId>mimepull</artifactId>
<version>1.9</version>
</dependency>
</dependencies>

總結是Jersey libraries跟mimepull的版本必須要相同. 不然會出現Missing Dependency的問題. 所以即使Jersey libraries已有1.14版, 這裡必須使用1.9版, 因為mimepull只有到1.9版.

完全沒想到missing dependency會跟版本有相關性.

希望這篇文章可以幫到需要的朋友們.

Original Tutorial Site:
http://www.mkyong.com/webservices/jax-rs/file-upload-example-in-jersey/

Sincerely,
Danil