NuoDB on Red Hat OpenShift

A few years ago, Red Hat introduced OpenShift, which provides a Platform as a Service (PaaS). There are a number of flavors of OpenShift from Origin, an open-source community project, to a fully managed online service. OpenShift allows developers to focus on incremental application changes while not having to administer the infrastructure. OpenShift provides a continuous integration framework where it slurps your GitHub repo into a docker container with the proper language engine and lets Kubernetes orchestrate the deployment of the container pods to various host nodes. That being said, NuoDB is not an application that can be deployed via the OpenShift web console but can be deployed as part of the PaaS. In this blog, I’ll walk you through standing up your own OpenShift Origin cluster on Centos and pull in NuoDB containers as part of the platform.

OpenShift Configuration

For this blog, I’m only launching a single OpenShift node on an AWS EC2 instance. This node contains both the “Master” and “Node” services. There are a number of ways to deploy an OpenShift cluster but I found the steps in this link to be the simplest and quickest to get a cluster up and running. Use the first option, “oc cluster up”. This will deploy the necessary OpenShift containers, create a default oc user account, and create a new project.

When running the cluster up command, you should get the following results:

[root@ip-172-31-17-249 ~]# oc cluster up
-- Checking OpenShift client ... OK
-- Checking Docker client ... OK
-- Checking Docker version ... OK
-- Checking for existing OpenShift container ... OK
-- Checking for openshift/origin:v1.4.1 image ... 
   Pulling image openshift/origin:v1.4.1
   Pulled 0/3 layers, 3% complete
   Pulled 1/3 layers, 34% complete
   Pulled 2/3 layers, 79% complete
   Pulled 3/3 layers, 100% complete
   Extracting
   Image pull complete
-- Checking Docker daemon configuration ... OK
-- Checking for available ports ... OK
-- Checking type of volume mount ... 
   Using nsenter mounter for OpenShift volumes
-- Creating host directories ... OK
-- Finding server IP ... 
   Using 172.31.17.249 as the server IP
-- Starting OpenShift container ... 
   Creating initial OpenShift configuration
   Starting OpenShift using container 'origin'
   Waiting for API server to start listening
   OpenShift server started
-- Adding default OAuthClient redirect URIs ... OK
-- Installing registry ... OK
-- Installing router ... OK
-- Importing image streams ... OK
-- Importing templates ... OK
-- Login to server ... OK
-- Creating initial project "myproject" ... OK
-- Removing temporary directory ... OK
-- Server Information ... 
   OpenShift server started.
   The server is accessible via web console at:
       https://172.31.17.249:8443

   You are logged in as:
       User:     developer
       Password: developer

   To login as administrator:
       oc login -u system:admin      

OpenShift should have launched your necessary containers. You can verify this by running the Docker command “docker ps” which will return the following running containers. If any of them have failed, you can view the log output using “docker logs” or restart the container with “docker start.”

        [root@ip-172-31-17-249 ~]# docker ps
CONTAINER ID        IMAGE                           COMMAND                  CREATED   STATUS     PORTS   NAMES
22139909dd41        origin-docker-registry:v1.4.1   "/bin/sh -c 'DOCKER_R"   2 mins    up 2 mins          k8s_registry...
8a9cf9785d91        origin-haproxy-router:v1.4.1    "/usr/bin/openshift-r"   2 mins    up 2 mins          k8s_router...
a24548fc86ff        origin-pod:v1.4.1               "/pod"                   2 mins    up 2 mins          k8s_POD...
3c2e9b053ed9        origin-pod:v1.4.1               "/pod"                   2 mins    up 2 mins          k8s_POD...
08e31f593cbd        origin:v1.4.1                   "/usr/bin/openshift s"   2 mins    up 2 mins          origin      

Log Into OpenShift

The “master” Docker container is running the OpenShift web interface. You can log into the interface by browsing to the host’s IP address on port 8443. For example:

https://172.31.17.249:8443

The username and password “cluster up” created is ‘developer/developer.’ To deploy NuoDB, we’ll need to log into the CLI.  This can be done locally or from another box running the oc client tools. The following example is logging in locally:

        [root@ip-172-31-17-249 ~]# oc login -u developer -p developer
Login successful.

You have one project on this server: "myproject"

Using project "myproject".      

As I mentioned before, the “oc cluster up” command has already created a new project for you. Checking the status of our project will show that no applications have yet been created.

        [root@ip-172-31-17-249 ~]# oc status
In project My Project (myproject) on server https://172.31.17.249:8443

You have no services, deployment configs, or build configs.
Run 'oc new-app' to create an application.      

We are going to remove the default project and create our own:

        [root@ip-172-31-17-249 ~]# oc delete project myproject
project "myproject" deleted

[root@ip-172-31-17-249 ~]# oc new-project nuodb-demo
Now using project "nuodb-demo" on server "https://172.31.17.249:8443".

You can add applications to this project with the 'new-app' command. For example, try:

    oc new-app centos/ruby-22-centos7~https://github.com/openshift/ruby-ex.git

to build a new example application in Ruby.      

We need to make one more OpenShift system configuration. NuoDB runs as the root within our Docker containers so we’ll need the policy to allow containers to run as root. First, elevate your oc permissions:

        [root@ip-172-31-17-249 ~]# oc login -u system:admin
Logged into "https://172.31.17.249:8443" as "system:admin" using existing credentials.

You have access to the following projects and can switch between them with 'oc project <projectname>':

    default
    kube-system
  * nuodb-demo
    openshift
    openshift-infra

Using project "nuodb-demo".      

Then set the policy to allow containers to run as any user:

        [root@ip-172-31-17-249 ~]# oc adm policy add-scc-to-user anyuid -z default      

We are now ready to deploy NuoDB containers.

NuoDB Deployment

Before we deploy our NuoDB containers, we need to disable transparent huge page for the NuoDB storage manager to operate correctly. Execute the following in a shell file on the server:

        #!/bin/sh

disable_thp() {

    echo "Checking to see if THP needs to be disabled..."
    THP_BASE=/sys/kernel/mm/transparent_hugepage
    if [ ! -f ${THP_BASE}/enabled ]; then
        THP_BASE=/sys/kernel/mm/redhat_transparent_hugepage
        if [ ! -f ${THP_BASE}/enabled ] ; then
            die "unable to find THP enable file"
        fi
    fi
    THP_ENABLED=${THP_BASE}/enabled
    THP_DEFRAG=${THP_BASE}/defrag 
    THP_WAS_ENABLED=false
    grep "\[always\]" ${THP_ENABLED} 
    if [ $? -eq 0 ]; then
        THP_WAS_ENABLED=true
        echo "THP is enabled on this machine - this script will temporarily disable it"
        echo madvise> ${THP_ENABLED}
        if [ $? -ne 0 ]; then

            die "Cannot automatically disable THP on a readoly file system - See http://doc.nuodb.com/display/doc/Linux+Installation#LinuxInstallation-DisablingTHPonRead-OnlyFileSystems"
        else
            echo madvise> ${THP_DEFRAG}
        fi 
    else
        echo "THP is not enabled on this machine - No action taken"
    fi
}

disable_thp      

I’ve made deploying NuoDB as simple as possible. The following command will create a temporary OpenShift application. This deployment container will pull a NuoDB CE image from docker.io registry and deploy a Broker, SM (storage manager), and TE (transaction engine).  

        oc new-app docker.io/nuodbopenshift/nuodb-deployer:latest \
     --name nuodb-deployer \
     -e "OC_ADDRESS=<host_IPAddress>" \
     -e "USERNAME=developer" \
     -e "PASSWORD=developer"       

The environment variables are required for the deployment container to establish a CLI session with OpenShift. OC_ADDRESS is the IP address of your OpenShift instance.

In the OpenShift web console, you should now see three applications under “Overview.” The console will allow you to scale up or down your pods to match load demands. The version of NuoDB is a community edition and has some scaling restrictions but will be sufficient for testing your applications against.

You now have NuoDB up and running on OpenShift. The first database has been created for you called ‘testdb’ with a username and password of ‘dba/dba’. Your application’s connection string will point to the broker’s IP address on port 48004. To get the IP address of your broker use the following command: oc describe pod broker | grep IP: | awk '{print $2}').