The ability to make use of VMware Site Recovery Manager has allowed RBKC to squeeze more value out of its Clariion CX500 arrays, which support an environment in which nearly half of the servers are virtualised. The EMC arrays are nearing end of life after expanding from 2 TB at implementation to approximately 20 TB now, but FalconStor storage virtualisation means the borough will be able to add arrays from any vendor to expand capacity.
The west London local authority has 3,500 employees serving 178,600 residents. It has virtualised approximately 42% of its servers onto six VMware ESX nodes and 135 virtual machines (VMs). Data storage capacity is on two EMC Clariion CX500 storage-area networks (SANs): one at the town hall and a second DR site at a location one mile away. FalconStor manages data replication of logical unit numbers (LUNs) between the two sites. Capacity has been reached on the EMC array at the primary site, and the borough's IT team will soon begin the process of procuring more SAN capacity.
RBKC originally implemented FalconStor appliances (then known as IPStor, now called NSS) alongside each EMC array three years ago. The product allowed the borough to virtualise storage between the two sites, replicate data across the two arrays, presented Fibre Channel and iSCSI LUNs to virtual servers, and provided NAS file-based access.
With the move to FalconStor NSS 6.0, the borough can now use the automated recovery of backed up virtual servers enabled by VMware Site Recovery Manager. VMware Site Recovery Manager allows users to specify policies and procedures in VMware vCenter for automated recovery of virtual server clusters. Upon the invocation of a DR plan, this allows storage administrators to press a button to set the process in motion.
"We're really pleased we can now use SRM," Goodall said. "Until this version FalconStor didn't have the APIs to work with it, so we had this feature in VMware but couldn't use it. Now we can. I also really like the 'bubble' test environment in SRM that allows us to run tests on recovering from a disaster."
RBKC chooses FalconStor based on cost and functionality
Goodall's team originally chose FalconStor in conjunction with EMC arrays based on cost and functionality. "It was cheaper to set things up with FalconStor than use all EMC [products]. We wanted two SANs, replication between the two sites and NAS, and the EMC/Falconstor solution came in cheaper," he said.
RBKC also looked at Dell/EMC, StorageTek and a pure EMC product. The Dell/EMC offering was rejected as too expensive; StorageTek, fell out of favor when it was then being acquired by Sun Microsystems and perceived as having an uncertain future; and a pure EMC solution from Computacenter lost out, Goodall said, when it was determined that seven pieces of EMC software would be required.
When asked if there was anything FalconStor could do better, Goodall mentioned snapshot handling. "We've been looking at procuring more SAN capacity and we've taken a look at NetApp. One thing we like is the way it automatically presents snapshot data into folders that can be accessed by end users. If we wanted to do that in FalconStor, it would take a lot of administrator time and effort."