Manual Installation Guide

Introduction

This document explains how to manually install StorReduce on a server.

Please note that wherever possible we recommend using the StorReduce Virtual Appliance an VMWare virtual machine image, which comes the StorReduce server pre-installed and ready to run, or install StorReduce from the AWS, Azure or Google Cloud Marketplaces.

Before You Start

Please contact StorReduce support and ask for the install-from-yum.sh script. This script will install the StorReduce YUM repository and then use that to install the StorReduce RPMs.

Requirements

Linux Distribution

The StorReduce server runs on Red Hat based Linux distributions (e.g., RHEL, CentOS, Fedora, AMI Linux, …). We support manual installation on Red Hat 6.6+ and CentOS 6.6+.

Hardware Requirements

Please see the Hardware Requirements section of the On-Premises Installation Guide.

Disk Configuration

As discussed in the On-Premises Installation Guide, StorReduce requires two disks to store its deduplication index. At least one of these disks should be an SSD. For larger work loads (more than 100TB of logical, pre-deduplication, data) both disks should be SSDs if possible.

These disks should be formatted with the ext2 filesystem1 and mounted at /mnt/srdb and /mnt/srkeydb.

If the high iops disk is /dev/sdX if the standard disk is /dev/sdY then run then run these commands (as root) to format the disks:

mkfs -t ext2 /dev/sdX
mkfs -t ext2 /dev/sdY

then make mount points like this:

mkdir -p /mnt/srdb
mkdir -p /mnt/srkeydb

To ensure the disks are mounted when the server restarts, add entries like this to /etc/fstab:2

/dev/sdX    /mnt/srdb       ext2    defaults,noatime  0  2
/dev/sdY    /mnt/srkeydb    ext2    defaults,noatime  0  2

Finally, mount the disks for the first time like this:

mount /mnt/srdb
mount /mnt/srkeydb

Firewall

The StorReduce server uses TCP ports 80, 443 and 8080 so you will need to configure the firewall to allow connections from those ports.

Manual StorReduce Installation

To manually install StorReduce:

  1. Copy the install-from-yum.sh script that you obtained from StorReduce support onto your Linux server.

  2. Ensure that the /mnt/srdb and /mnt/srkeydb directories are mounted (see above).

  3. Run the script as root:

    chmod +x ./install-from-yum.sh
    ./install-from-yum.sh
    
  4. Start supervisord and it will start StorReduce:

    service supervisord start
    

You should now be able to access the StorReduce server via a URL like this:

https://SERVER.IP.ADDRESS:8080

To configure the server see the Configure your StorReduce Server section of Quick Start Guide: On-Premises Server.

StorReduce can be stopped, started and restarted using supervisorctl:

supervisorctl stop storreduce
supervisorctl start storreduce
supervisorctl restart storreduce

What was Installed?

The install-from-yum.sh script installed:

  1. Configuration for the Extra Packages for Enterprise Linux (EPEL) YUM repository. EPEL is a collection of useful packages for Red Hat and compatible Linux distributions maintained by the Fedora project.

  2. curl which is used by some of the scripts installed later.

  3. pip a package management system for the Python language, some of the support scripts and other tools installed later use Python packages.

  4. Configuration for the StorReduce YUM repository.

  5. aws-cli a command line program that, among other things, can be used to copy files in and out of StorReduce. See Using the AWS Command Line Interface to access StorReduce for more information.

  6. supervisord a process control system. It starts and stops the StorReduce Server, and manages its logs.

  7. prometheus a statistics server which is used to record statistics about the StorReduce Server and produce graphs from those statistics.

  8. The StorReduce server binary /usr/bin/storreduce and it’s start script /usr/bin/start-storreduce.

  9. Init scripts to start supervisord and storreduce when the server is restarted.

You can change the configuration of StorReduce by altering the command line flags passed into the binary. To see the available flags, run storreduce --help. To change the flags alter /usr/bin/start-storreduce, but please take a backup copy of your changes since they may be overwritten when StorReduce is overwritten.

Upgrading

Since StorReduce is installed using YUM, it can be upgraded using YUM. Just run:

yum update 'storreduce*' 

To install the latest updates.


  1. We use the ext2 filesystem because it offers the best performance with memory mapped databases.

    [return]
  2. The noatime option reduces the number of writes into filesystem by disabling update to the last accessed time attribute on the files in the filesystem.

    [return]