Alamy is the world\u2019s largest online photo agency for stock, video and live news. Additionally, it has one of the fastest websites in the industry, gathering over 30,000 individual photographers and about 600 picture agencies worldwide. Every day over 30,000 new images are being added to their already highly developed collection.<\/p>\n
What was the challenge?<\/strong><\/h4>\nAn important fact is that Alamy\u2019s Head Office operations are supported by a diverse range of servers. Their infrastructure consisted of a single Open-E DSS V6 license providing iSCSI storage to two ESXi servers. The third ESXi server hosted a virtualised Exchange server. It turned out that all hardware was approaching EOL (end of life). As performance issues with Exchange were occurring on a regular basis, despite multiple dedicated disk arrays, the company wanted to increase their resiliency, mainly by eliminating single points of failure. Unfortunately VMware\u2019s licensing requirements for live migration (Vcenter), as well as achieving true redundancy of shared storage by means of traditional SAN solutions, were far beyond Alamy\u2019s budget.<\/p>\n
Solution<\/strong><\/h4>\nAlamy\u2019s IT Team decided to try out a solution leveraging a high availability hypervisor platform, Hyper-V 2012 clustering, and Open-E DSS V7<\/a> with Active-Active iSCSI Failover<\/a> to provide high availability iSCSI storage. First, they created a test system with four existing test-lab servers. Then the system was evaluated with different configuration options and finally stress tested \u2013 to make sure the new environment performs well and that the failover of Open-E and Hyper-V works as expected.<\/p>\n![\"Open-E](\"https:\/\/www.open-e.com\/blog\/wp-content\/uploads\/2015\/03\/Open-E-Case-Study-Alamy.jpg\")
<\/a><\/p>\nWhen it came to hardware, Alamy used 2 x Dell PowerEdge R620 servers, each with 2 x 10core processors and 128GB ram as Hyper-V servers, and 2 x Dell PowerEdge R720XD servers, each with 12 x 2TB SATA drives as Open-E servers. Then, the existing servers were migrated from ESX to Hyper-V.<\/p>\n
Challenge accomplished!<\/strong><\/h4>\n\n“The active-active iSCSI replication in Open-E DSS V7 provides load-balancing of storage, redundancy against hardware failure, and the ability to install hardware and OS updates without VM downtime which were the strongest arguments for our decision. Additionally, our Exchange performance issues have completely disappeared.<\/em><\/p>\nMigrating from ESX to Hyper-V means that both failover clustering and live migration are supported without a large investment in licenses, providing load balancing, redundancy against hardware failure, and the ability to install hardware and OS updates without VM downtime. With MPIO on Open-E DSS V7 and NIC bonding with MPIO on Hyper-V, even the failure of a NIC causes no downtime. An actual hardware failure demonstrated that the project goal of resilience had been achieved.”<\/em><\/p>\nMark Jones \u2013 Network Administrator at Alamy<\/div>\n<\/div>\n <\/p>\n
If you would like to to find out more, read the full case study >><\/a><\/p>\n
Alamy\u2019s IT Team decided to try out a solution leveraging a high availability hypervisor platform, Hyper-V 2012 clustering, and Open-E DSS V7<\/a> with Active-Active iSCSI Failover<\/a> to provide high availability iSCSI storage. First, they created a test system with four existing test-lab servers. Then the system was evaluated with different configuration options and finally stress tested \u2013 to make sure the new environment performs well and that the failover of Open-E and Hyper-V works as expected.<\/p>\n When it came to hardware, Alamy used 2 x Dell PowerEdge R620 servers, each with 2 x 10core processors and 128GB ram as Hyper-V servers, and 2 x Dell PowerEdge R720XD servers, each with 12 x 2TB SATA drives as Open-E servers. Then, the existing servers were migrated from ESX to Hyper-V.<\/p>\n “The active-active iSCSI replication in Open-E DSS V7 provides load-balancing of storage, redundancy against hardware failure, and the ability to install hardware and OS updates without VM downtime which were the strongest arguments for our decision. Additionally, our Exchange performance issues have completely disappeared.<\/em><\/p>\n Migrating from ESX to Hyper-V means that both failover clustering and live migration are supported without a large investment in licenses, providing load balancing, redundancy against hardware failure, and the ability to install hardware and OS updates without VM downtime. With MPIO on Open-E DSS V7 and NIC bonding with MPIO on Hyper-V, even the failure of a NIC causes no downtime. An actual hardware failure demonstrated that the project goal of resilience had been achieved.”<\/em><\/p>\n <\/p>\n If you would like to to find out more, read the full case study >><\/a><\/p>\n<\/a><\/p>\nChallenge accomplished!<\/strong><\/h4>\n