Posted by Ashish Shah on Wed, 2011-11-02 10:09
I met with a prospective customer last week that happens to be one of the largest Internet portals in the world, with tens of data centers across the US as well as in Europe and Asia. One would think that with such an expansive infrastructure, some form of WAN optimization solution would already be in place across data centers. When I asked which WAN optimization solution they had deployed to date, the chief network architect laughed at me and said: “We have over 100Gbps connectivity between most of our data centers. There’s no way we would deploy hundreds of 1Gbps-capable devices at each site!” He then went on to tell us his previous experience with other WAN optimization solutions, and how he bought a pair from a vendor and was using them to decorate the storage shelves in his lab.
Posted by Ashish Shah on Wed, 2011-10-26 01:55
A large healthcare firm we recently talked to had an upcoming data migration project, where the IT team planned on moving 800TB of data from a to-be-decommissioned data center to a larger data center 1,100 miles away. The WAN connectivity between the two sites consisted of 2x1Gbps MPLS links. The storage and network team concluded that the migration would take four months to complete, making it operationally cost-prohibitive. The network team considered leasing a 10Gbps WAN link, but came to the conclusion that the significantly higher WAN expenditure would in no way guarantee faster migration. Interestingly enough, one-time or periodic migrations are becoming quite common for medium to large enterprises.
