For past couple of months DCOS is has gained its popularity exponentially. DCOS has updated from version 1.6-1.10 with tremendous integration with different applications. It supports different tools from Hadoop technology to make the process fast and efficient to manage all of them from a single console.
To install HDFS over DCOS we must have healthy cluster minimum of 11 nodes and 1 external node to install DCOS CLI. However for better performance/production environment it is better to add at least 25 node with handsome configuration.
Machine
|
Core
|
RAM
|
HDD
|
BootstrapNode
|
>= 8
|
>= 16
|
40GB
|
MesosMaster1
|
>= 8
|
>= 12
|
40GB
|
MesosMaster2
|
>= 8
|
>= 12
|
40GB
|
MesosMaster3
|
>= 8
|
>= 12
|
40GB
|
MesosAgent1
|
>= 6
|
>= 6
|
40GB
|
MesosAgent2
|
>= 6
|
>= 6
|
40GB
|
MesosAgent3
|
>= 6
|
>= 6
|
40GB
|
MesosAgent4
|
>= 6
|
>= 6
|
40GB
|
MesosAgent5
|
>= 6
|
>= 6
|
40GB
|
MesosAgent6
|
>= 6
|
>= 6
|
40GB
|
MesosPublic
|
>= 6
|
>= 6
|
40GB
|
With the above number of Machines, DCOS cluster we will have 1 Bootstrap Node from where we install all the packages on DCOS Master Node and 6 PrivateAgent nodes and 1 PublicAgent respectively. However we can increase the number of datanode as per our requirement if we have enough number of Mesos agent in our DCOS cluster.
DCOS can be install either on-premise or over Cloud Platfrom (AWS/Azure) on certain OS like Centos/RHEL/CoreOS etc. The steps given below are perfectly working on Centos7.2 for both on-premise and AWS Cloud.
Pre-requisite for DCOS Intallation
There are certain pre-requisite steps that needs to be followed on either of the machines listed in the above table as below:
1. Login as a root and create a new group and user respectively. Add UID (1000 or above to a user) and set password for the mesos user
sudo su –
groupadd -g 1001 mesos
useradd -g mesos -u 1001 mesos
passwd <mesos>
2. Edit the config file(Via Root) and reboot the system to reflect
vi /etc/selinux/config
%s/SELINUX=enforcing/SELINUX=disabled/g
:wq!
3. Set the hostname as below
sudo hostnamectl set-hostname mesos-agent1.machine.com --static
4. Edit network file and add the below line
vi /etc/sysconfig/network
HOSTNAME= mesosmaster (same name as in the step3)
NETWORKING_IPV6=no
IPV6INIT=no
:wq!
Note: In case of creating cluster in EC2 environment please add the complete private domain name for host name
E.g.: If ip-192.X.X.X.internal private domain then assigned ip-192.X.X.X as a hostname variable
5. Login as root again and edit the sudoers file
su – root
vi /etc/sudoers
#Search for “wheel” and add the below line
mesos ALL=(ALL) NOPASSWD:ALL
:wq!
6. Set the Ulimit Value
ulimit –n 32768
ulimit –u 60000
7. Edit the limits.conf file
vi /etc/security/limits.conf
#Append the below line
mesos hard nofile 65536
mesos soft nofile 65536
mesos hard nproc 65536
mesos soft nproc 65536
:wq!
8. Edit the file /etc/sysctl.conf file and add the below line at the end
vi /etc/sysctl.conf
kernel.pid_max = 4194303
net.ipv4.ip_local_port_range = 1024 64000
net.ipv6.conf.all.disable_ipv6 = 1
:wq!
9. Disable firewall in all the machine
sudo systemctl stop firewalld && sudo systemctl disable firewalld
10. Disable the Iptables
sudo chkconfig iptables off
sudo chkconfig ip6tables off
11. Copy the IP of each machine in host file on every machine in the DCOS cluster and execute the below command with updated IP of your cluster
sudo tee -a /etc/hosts << -'EOF'
10.5.3.80 bootstrap.machine.com
10.5.3.10 mesos-master3.machine.com
10.5.3.11 mesos-master2.machine.com
10.5.3.12 mesos-master1.machine.com
10.5.3.20 mesos-agent6.machine.com
10.5.3.21 mesos-agent5.machine.com
10.5.3.22 mesos-agent4.machine.com
10.5.3.23 mesos-agent3.machine.com
10.5.3.24 mesos-agent2.machine.com
10.5.3.25 mesos-agent1.machine.com
10.5.2.40 mesos-public.machine.com
EOF
Note: In case of ec2 environment we need to add the private domain name and internal IP address as below, complete below line should be added
192.X.X.X ip-192.X.X.X.internal ip-192.X.X.X
12. ssh key based authentication is required so that each of the machine can communicate with each other without password
su – mesos
ssh-keygen –t rsa (Press Enter for default )
#Press enter again for default passphrase
cat /home/mesos/.ssh/id_rsa.pub >> /home/mesos/.ssh/authorized_keys
ssh localhost (Enter yes)
exit (connection should be closed)
Note: Copy the contents from id_rsa, id_rsa_pub & authorized_keys into respective files on each machine at same user/path with same privilege to ensure each machine can talk to other without password. Reboot all the machines to reflect the changes made in above steps.
13. Update the yum repo and install certain packages on each machine in the cluser
sudo tee /etc/yum.repos.d/docker.repo <<-'EOF'
[dockerrepo]
name=Docker Repository
baseurl=https://yum.dockerproject.org/repo/main/centos/$releasever/
enabled=1
gpgcheck=1
gpgkey=https://yum.dockerproject.org/gpg
EOF
sudo mkdir -p /etc/systemd/system/docker.service.d && sudo tee /etc/systemd/system/docker.service.d/override.conf <<- EOF
[Service]
ExecStart=
ExecStart=/usr/bin/dockerd --storage-driver=overlay
EOF
sudo yum upgrade --assumeyes --tolerant && sudo yum update --assumeyes
sudo yum install -y mlocate.x86_64 tree screen net-tools.x86_64 bind-utils
sudo yum -y install epel-release
sudo yum -y install build-essential python-dev libcurl4-nss-dev libsasl2-dev libsasl2-modules maven libapr1-dev libsvn-dev curl wget ntp htop;
#Install Docker
sudo yum install -y docker-engine-1.13.1 docker-engine-selinux-1.13.1
#Start Services
sudo systemctl start docker && sudo systemctl enable docker && sudo service ntpd start && sudo sysctl -w net.bridge.bridge-nf-call-iptables=1 && sudo sysctl -w net.bridge.bridge-nf-call-ip6tables=1
# Update the java path in bash profile
vi ~/.bashrc
readlink -f /usr/bin/java | sed "s:bin/java::"
export JAVA_HOME=$(readlink -f /usr/bin/java | sed "s:bin/java::")
wq!
Note: Make sure Java version should be >=1.7
Additional Steps On BootStrap Node
Login to the Bootstrap node and follow the below steps to launch DCOS
1. Goto home directory and create a genconf directory and update the config file with IP as below
mkdir -p genconf && tee genconf/config.yaml <<-'EOF'
agent_list:
- 10.5.3.20
- 10.5.3.21
- 10.5.3.22
- 10.5.3.23
- 10.5.3.24
- 10.5.3.25
bootstrap_url: file:///opt/dcos_install_tmp
cluster_name: DEMO-DCOS
exhibitor_storage_backend: static
ip_detect_public_filename: genconf/ip-detect-public
ip_detect_path: genconf/ip-detect
master_discovery: static
master_list:
- 10.5.3.10
- 10.5.3.11
- 10.5.3.12
public_agent_list:
- 10.5.2.22
process_timeout: 10000
resolvers:
- 8.8.8.8
- 8.8.4.4
ssh_key_path: genconf/ssh_key
ssh_port: 22
ssh_user: mesos
EOF
Note: This is a minimum configuration, we can do more customization as per the requirement
2. Create a script ip-detect as below
tee genconf/ip-detect <<-'EOF'
#!/usr/bin/env bash
ip addr show |grep "inet " |grep -v 127.0.0. |head -1|cut -d" " -f6|cut -d/ -f1
EOF
3. Download the script in home directory
curl -O https://downloads.dcos.io/dcos/stable/dcos_generate_config.sh
4. Execute the Below command
sudo bash dcos_generate_config.sh --genconf &&
sudo bash dcos_generate_config.sh --install-prereqs &&
sudo bash dcos_generate_config.sh --preflight &&
sudo bash dcos_generate_config.sh --deploy &&
sudo bash dcos_generate_config.sh --postflight;
Below message should be displayed upon successful completion.
====>
EXECUTING INSTALL PREREQUISITES
====> START
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> STAGE
install_prereqs
====> OUTPUT
FOR install_prereqs
====> ACTION
run_preflight COMPLETE
====>
SUMMARY FOR run_preflight
10 out of 10
hosts successfully completed run_preflight stage.
====> END OF
SUMMARY FOR run_preflight
====>
EXECUTING DC/OS INSTALLATION
====> START
install_dcos
====> ACTION
run_postflight COMPLETE
====>
SUMMARY FOR run_postflight
10 out of 10
hosts successfully completed run_postflight stage.
====> END OF
SUMMARY FOR run_postflight
4. Go to the browser and launch the below URL
http://<master-public-ip>:8181/exhibitor/v1/ui/index.html
http://<public-master-ip>/
http://<public-master-ip>:8080
Login to the DCOS console by selecting either of the account type and Enjoy the DCOS Cluster!!!
Hope you have enjoyed the post and able to setup the environment without any issue. For any issue/troubleshoot please feel free to approach, always ready to help. Can assist to troubleshoot any issue.
For custom setup to expand the DCOS existing cluster, stay tuned with my blog.
For custom setup to expand the DCOS existing cluster, stay tuned with my blog.
No comments:
Post a Comment