Quick-Start Guide: Deploying your Cloudian Hyperstore Hybrid

Storage Service

Contents Intended Audience .......................................................................................................................... 2

Cloudian Node Requirements: ........................................................................................................ 2

Preparing cluster node hardware ................................................................................................... 3

Installing the Cloudian HyperStore Cluster ..................................................................................... 5

Completing Your Cloudian HyperStore Setup................................................................................. 9

AWS Proxy Server ......................................................................................................................... 12

Accessing the AWS Proxy Console ................................................................................................ 18

Intended Audience The purpose of this document is to help a new user deploy a 3-node Cloudian storage cluster in your datacenter for use with the Cloudian HyperStore Hybrid Cloud Service from AWS Marketplace. This document is a quick start guide and does not replace the existing Cloudian HyperStore Administrator’s Guide or the Cloudian HyperStore Installation Guide. It is recommended to refer to the Cloudian Administrator’s Guide and the Installation Guide for additional details and larger configurations than 3nodes.

Cloudian Node Requirements: 3 Intel compatible servers each with the following minimum configuration

Per Server Minimum Hardware Configuration o Processor: 1 CPU, 8 cores, 2.4GHz o Memory: 32GB o Disk: 12 x 2TB HDD, 2 x 160GB SSD (12 drives for data, 2 drives for OS/Metadata) o RAID: RAID-1 recommended for the OS/Metadata, JBOD for the Data Drives o Network: 2x10GbE Port

Note. Example Cloudian HyperStore Setup In the remainder of this document, we will be installing a 3-node Cloudian HyperStore cluster. Our servers are:

cloudiannode1 IP address: 192.168.2.41

cloudiannode2 IP address: 192.168.2.42

cloudiannode3 IP address: 192.168.2.43 We are installing our setup using our AD and DNS server. The domain is: dom.local

Preparing cluster node hardware 1. Use the Cloudian HyperStore ISO for OS installation on all 3 nodes. This will install CentOS on your

new servers along with the HyperStore software binary files. An ISO is a bootable image which will deploy the OS software, related packages and Cloudian software onto a node. Follow these steps to install CentOS from the HyperStore Appliance ISO file:

Burn the HyperStore Appliance ISO file on to either a CD or a DVD. Be sure to use a "burn image" operation, not a simple copy operation.

Insert the CD or DVD into the host machine.

Start the host machine (or restart it if it’s already running) and use your hardware-specific hot key to enter the BIOS setup utility.

Use the BIOS setup utility to change the machine’s boot order so that it boots first from the CD/DVD drive. When you save your changes, the machine will restart and boot from the ISO file. After a few moments, the CentOS Installer welcome screen will briefly appear.

Select the first option in the welcome menu, "Install or upgrade an existing system", and press enter. If you do not make a selection within 5 seconds, the "Install or upgrade an existing system" option automatically executes.

o The console will display progress information as the packages are installed. There will be a pause for a couple minutes after the last package is installed, and then post-installation scripts will run. Altogether the system installation will typically take less than 10 minutes to complete.

After the install has completed, the CD/DVD tray ejects and the system begins to reboot. Remove the CD/DVD from the tray when it ejects so that the machine does not boot from the CD/DVD again.

CentOS 6.8 and HyperStore prerequisites are now installed on the host, and the HyperStore binaries and host set-up tool have been placed on the host.

2. Configure each node

a. Login to the node. The default root password is password. It is our best practice recommendation to update your root access for production environments, which we will do later in the process.

b. cd to /root/CloudianTools directory. c. configure_appliance.sh performs the following tasks, run through each step in order:

i. Change the default root password Although it is possible to configure a different root password for each node in the cluster, this would prevent the network validation tool running, so bear this in mind when setting passwords.

ii. Change time zone iii. Configure network

d. The interactive script will ask you to provide the information below. i. Hostname - Hostname of the machine. Do not use capital letters in the hostname,

and do not use localhost as the hostname.

ii. Network Interface - Network interface device to which these configurations will be

applied; for example, "eth0". iii. IP Address - IP (v4) address of the network interface being configured. This must be a

static IP address (not DHCP). Do not use the loopback address (127.0.0.1). Do not use IPv6.

iv. Netmask - Netmask for the subnet in which this host machine will be deployed. v. Default Gateway - IP address of the default gateway device for the subnet in which

this host machine will be deployed. vi. Primary DNS server - IP address of the DNS server that this host will use.

vii. Secondary DNS server - IP address of the secondary DNS server that this host will use, if any.

viii. Domain - Domain to which this host will belong. This is the fully qualified domain name (FQDN) of the host, minus the hostname segment. For example:

If the host’s hostname is hyperstore1 and its FQDN will be hyperstore1.enterprise.com, then specify enterprise.com as the domain.

If the host’s hostname is hyperstore1 and its FQDN will be hyperstore1.west.enterprise.com, then specify west.enterprise.com as the domain.

3. Format and mount available disks for Cloudian S3 data storage The script will automatically configure and mount available disks as ext4 file systems, for storage of HyperStore S3 objects. The mount points will be named as /cloudian1, /cloudian2, and so on. Console output will inform you of the progress for each disk as the script creates a partition, creates an ext4 file system, and mounts the file system.

If the node has 10 or more hard drives, then during the ISO install the two smallest hard drives were mirrored and used for the OS. Those drives will also store Cassandra data, Cassandra commit logs, and Redis data. The rest of the drives are considered available for the configure_appliance.sh script to mount as ext4 file systems for S3 object storage.

4. Reboot the node.

Rebooting is mandatory in order to apply the network configuration settings that you specified.

If necessary for your particular hardware, on reboot enter the BIOS utility and set the boot order back to booting first from the hard drive rather than the CD/DVD drive.

5. Repeat this process on every node in the HyperStore cluster. The same HyperStore ISO CD/DVD is

used for each host.

Installing the Cloudian HyperStore Cluster Under /root, you will find the CloudianPackages directory: 1. Run ./CloudianHyperStore-6.0.1.2.bin cloudian_xxxxxxxxxxxx.lic to extract the package content from

one of your nodes. This will node will be designated the Puppet master node. Note: cloudian_xxxxxxxxxxxx.lic is the license file provided to you by Cloudian and should be placed in the CloudianPackages directory before installation is started.

2. Create a copy of the sample-survey.csv file to survey.csv

3. Edit the new survey.csv file

In the survey.csv file, specify the region, the node name(s), IP address(s), DC, and RAC of your Cloudian HyperStore S3 cluster. In this example, our survey.csv file has the following updates for the 3-node cluster:

region2,cloudiannode1,192.168.2.41,DC1,RAC1

region2,cloudiannode2,192.168.2.42,DC1,RAC1

region2,cloudiannode3,192.168.2.43,DC1,RAC1 NOTE: You can specify an additional NIC on your x86 servers for internal cluster communication. If there are multiple NICs, the install script will ask you to specify the NIC for internal cluster communication.

Although it is possible to configure a different root password for each node in the cluster, this would prevent the network validation tool running, so bear this in mind when setting passwords.

4. Network Validation

From the /root/CloudianPackages/preinstall directory execute the preInstallCheck.sh script. This will provide an assessment of the node and network environment in order to ensure that the relevant

technical pre-requisites are in place to support the next stage of deployment. The script will output warnings and errors that may cause the software installation to fail if not rectified.

The preInstallCheck.sh script can be run with the parameters below;

The preInstallCheck.sh script provides a summary at the end of the run that outlines any environmental issues that has been detected. These should be resolved before continuing with the next installation phases.

Execute ./preInstallCheck.sh –c to run the test.

Please refer to the minimum node specification and the networking preparations section to ensure that the environment is configured as necessary.

Note 1: preInstallCheck.sh uses the same password to login into each node, so if different passwords have been used across nodes for root user, this will prevent this script running effectively. Note 2: preInstallCheck.sh warns about IP addresses of other nodes not being present in /etc/hosts file on each node. This will warn before the installation as that stage has not been completed yet. Running this script again after install should show these tests to pass.

WARN No IP address found for cloudian-node1 in /etc/hosts on 10.111.0.102...

WARN No IP address found for cloudian-node3 in /etc/hosts on 10.111.0.102...

WARN No IP address found for cloudian-node1 in /etc/hosts on 10.111.0.103...

WARN No IP address found for cloudian-node2 in /etc/hosts on 10.111.0.103...

Note 3: Not all validation checks have hints yet to be displayed in the summary report. Mainly, the firewall detection checks need explaining when failing (as could be caused by very slow network or otherwise). Note 4: Recommended RAM configuration for node is specified at 32GB, but script warns at 16GB. This should be taken into account and the real limit is 32GB RAM.

5. Run ./cloudianInstall.sh and select ‘Install Cloudian HyperStore’. When prompted, input the survey.csv

file name. Continue with the setup. NOTE: If deploying in a non-production environment, it is possible that you may not have an existing DNS server or be able to allocate the minimum resource requirements for your servers. You can run your Cloudian install with ./cloudianInstall.sh configure-dnsmasq. Cloudian HyperStore includes an open source domain resolution utility to resolve all HyperStore service endpoints

6. In the following screenshot, the information that we had provided in the survey.csv file is used in the

Cloudian HyperStore cluster configuration. NOTE: In this non-production setup we are using a DNS server for domain name resolution.

Notice that the default CMC is renamed to cmc-region2.dom.local. If you are using a DNS server, update your forward and reverse lookup zones to include the correct service names for your Cloudian HyperStore cluster.

7. Your Cloudian HyperStore S3 Cloud Storage is now up and running.

8. You have successfully deployed a 3-node Cloudian HyperStore S3 cluster.

Completing Your Cloudian HyperStore Setup Access your Cloudian Management Console (CMC) using the configured URL. The default System Admin group user id is admin and the default password is public. It is our best practice recommendation to update your default admin password in a production environment.

NOTE: If you do have a DNS server in your environment, your Windows/Linux client that will access your Cloudian HyperStore must have mappings of IP addresses to host names. For a non-production setup that does not have a DNS server, update your Windows/Linux client host file to include the name resolution of your Cloudian HyperStore cluster. For example, include the following in your hosts file:

The CMC

The s3-region.domain.com

The bucket.s3-region.domain.com (the bucket name resolution can be updated when a group and user bucket is created).

From the Cloudian HyperStore console, we will complete the Storage Policies, Group, and SMTP settings. Click on Storage Policy to configure a logical storage policy.

Our first storage policy will be created with the following settings:

DATA DISTRIBUTION SCHEME: Replicas within a Single Datacenter

NUMBER OF REPLICAS: 1

DATA CONSISTENCY LEVEL: Strong

METADATA CONSISTENCY LEVEL: Strong

Click on Save to complete the Storage Policy creation. The next step in our setup is to create groups and users. Click on Users and Groups in the Cloudian HyperStore Console and select + New Group in Manage Group. It is our recommendation to use the Cloudian ‘? Help’ page for additional information on the granular settings and actions that can be performed.

For example, you can set the Group QoS or User QoS Group Default Limits on each user group that is created. Select Save to complete your user group creation. In our follow-up step, we will create a user account from the Manage Users menu. Click on + New User to add a new user. Complete the adding a new user and select Save.

Click on More for additional parameters that you can update for your user. You have just successfully completed the entire process of setting up a 3-node Cloudian HyperStore cluster, created a user group and a user. Sign out of the CMC as the administrator. As a user, login with the account that was just created.

As the Cloudian HyperStore S3 cloud storage user, you can dynamically create your own bucket and start uploading files and objects to protect your data.

The setup of a Cloudian S3 cloud storage solution is really that easy. Your users can start uploading, sharing files, pictures and also use it as a scale-out object storage for applications and analytics that support S3.

AWS Proxy Server An AWS Proxy Server is required for monitoring purposes. Here are the steps to deploy your AWS Proxy Server and connect to your on-premise Cloudian HyperStore cluster. Step 1: You would have started an AWS instance from the product page.

Step 2: From the AWS console, keep track of the following information of your EC2 instance:

1. The Instance Public IP address. In this example, it is 54.200.9.207

2. The instance ID. In this example, it is i-2a8e5ab3

Step3: Go to the on-premise Cloudian Hyperstore cluster and follow these steps:

a. Login using SSH into the staging node [where you unpacked the Hyperstore software]

b. Modify one of the configuration files [we use puppet]. It is recommended to copy the original

mts.properties.erb file to mts.properties.erb.orig before editing the file.

vi /etc/cloudian-6.0.2-puppet/modules/cloudians3/templates/mts.properties.erb

c. Go to the end of the file and add the following:

awsmmsproxy.host= proxy host server public IP address

In this example, it is awsmmsproxy.host =54.209.9.207

save the changes to the mts.properties.erb file.

d. Go to the staging directory [the place you unpacked the Hyperstore software] and invoke the

installer script cloudianInstall.sh

e. Select option 2 (Cluster Management) from the menu

f. In Cluster Management, select option b (Push Configuration Settings to Cluster)

The Puppet master node will push the changes to all Cloudian HyperStore nodes in your cluster.

The process should complete successfully.

g. Once we have pushed the configuration, go back to the Cluster Management Menu and choose

options “c” (Manage Services).

h. In the next step, select option “5” (S3 Service)

i. Type in the command ‘restart’ and press Enter.

The S3 services should restart successfully on all Cloudian HyperStore nodes.

You have just successfully deployed an AWS Proxy Server and connect it to your Cloudian HyperStore cluster. You can use the AWS Proxy instance to view and monitor the on-premise Cloudian HyperStore.

Accessing the AWS Proxy Console You can access your AWS Proxy Server from a supported web browser using:

http://<PUBLIC_IP_ADDRESS_OF_AWS_PROXY_INSTANCE>:8888/console

o In this example, it is ‘http://54.209.9.207:8888/console’

o The login username is: admin

o The password is: Instance ID