Virtual Server Infrastructure

Introduction to Entity Data Virtual Servers

Entity Data has been providing Virtual Server solutions in Australia for more than six years and in that time we've learnt what it takes to design, maintain and support a mission critical virtual infrastructure. We're an official VMware Partner and one of just a few companies officially recognized by VMware as a Virtualisation Service Provider.

Brief Introduction to Virtualisation

Virtualisation is a technique of splitting up the physical resources of a computer system and creating multiple "Virtual Servers". Each Virtual Server is completely independent of the other instances and can run it's own Operating System and Applications. The nature of virtualisation allows for very fast migration, recovery and failover of virtual servers between hardware nodes.

The Virtualisation technology that we use

Our infrastructure is supported by the latest hypervisor technologies from VMware. We're currently running VMware vSphere 4.1 the worlds leading virtual platform for enterprise requirements. VMware vSphere ensures complete isolation of resources between virtual environments resulting in the most robust virtualisation solution for your applications.

Redundancy and High Availability

Our infrastructure is highly available on both computing and disk infrastructures. Virtual resources are implemented within a highly available HA cluster while the disk environment is protected in a highly available SAN. HA provides the industry-highest level of Virtual Server infrastructure availability by leveraging on a fully-redundant clustered/cloud platform. This ensures that in the event that a Virtual Server hardware node fails, the Virtual Server is then re-located and re-started on another hardware node. This service provides exceptional redundancy removing all risk of physical hardware failure. vMotion technology allows us to migrate virtual servers live between hardware nodes without any interruption to the guest virtual machine.

Load Balancing and Distributed Resource Scheduling (DRS)

DRS (distributed resource scheduling) extends on the HA capability above by adding the ability for the Virtual Server to dynamically move within the cluster allowing processing power to be balanced. This migration occurs as an online real-time vMotion of the Virtual Server ensuring no disruption to the Virtual Server. This feature ensures that processing power will always be available when it's needed by your applications.

Some Benefits of a Virtual Server solution with Entity Data

Fully isolated Virtual Server solution hosted on VMware vSphere technology.

State of the art high-availability and geographical redundancy.

Multi-homed Cisco network infrastructure supported 5 upstream ISP's.

Full Root/Administrator Access to your Virtual Server.

Powerful Control Panel with Console Access.

Powerful Traffic Statistics via Netflow.

Daily Backups onto LTO-4 TAPE.

24/7/365 Technical Support.

All about our Servers

The foundation of our virtual infrastructure are our Dell Poweredge® R710 and R610 servers consisting of Intel Xeon® E5645 6-Core CPU's, 64GB's ECC DDR3 Memory, and Redundant Power Supplies that are fed independently from separate PDU's/UPS's. Four NIC's are in use on each node (two NIC's carrying management and backup traffic and two carrying virtual machine VLAN's) trunked across multiple switches for redundancy.

New hardware nodes being installed into Entity Data cabinets, March 2011

All about Storage and our Storage Area Network (SAN)

One of the most critically important design elements of a virtual environment is how storage is presented to virtual servers. Entity Data use only RAID-10 15,000 RPM Serial Attached SCSI disks for hosting virtual machines. Our Storage Area Network uses EqualLogic PS6010 iSCSI storage arrays connected to hardware nodes in a multi-pathed 10GbE configuration. Our iSCSI traffic has been completely isolated onto separate layer-2 switches for optimal performance. We can confidently say that we'll provide the fastest disk I/O possible by today's standards or your money back.

Network connectivity and our Data Centre's

What does it mean for you to host with an enterprise class network? It means speed, power, reliability, and 100% uptime! Our Enterprise-ISP network routes traffic to its destination using Border Gateway Protocol (BGP). BGP is an advanced network protocol that chooses how to send information through the Internet based on the fewest number of available network hops. We maintain our own redundant network infrastructure, including routers, switches and firewall's. Entity Data have a direct relationship with APNIC and use our own IP Address space and AS Numbers. We currently peer five network providers including; AAPT, Internode, Uecomm, Nextgen & PIPE Networks. Having fibre ethernet links to five upstream ISP's means we can guarantee 99.999% availability of our network.

Console access and web portals

Entity Data provide true out-of-bound console access to customers. This means that in the event you need to watch your server boot or need to fix a network misconfiguration you can simply login via the web control panel.

How we handle backups

Despite countless redundancies at the virtualisation level we still recognise the importance of backups. We do backups using Backup Exec 2010 R2: Virtual Infrastructure Edition which connects to VMware via the vStorage API. This allows for agent-less backups of customer virtual machines over the management network to prevent degradation of network performance. Backups are performed daily onto LTO-4 TAPE and allow for granular file restores or complete image/instance restores.

Traffic statistics via netflow

Customers have access to a proprietary web portal and can monitor incoming and outgoing data traffic from their server live, 24 hours per day. Email alerts are also sent when 80% and 100% of quota are reached.