Citrix has released software designed to speed up traffic between individual computers and corporate sites using its WAN acceleration appliances at only one end of the connection.
Until now, WANScaler devices talked only to each other to speed up traffic across WAN links, requiring a WANScaler at each end of the connection.
The new WANScaler client software makes it possible for individuals working remotely to gain the same benefits, Citrix says. Similar software is on the roadmap for Juniper Networks’ WAN acceleration products.
Optimisation gear can speed up applications over WAN links, making server consolidation more feasible because performance doesn’t degrade as much when applications are accessed from remote server farms rather than from local servers.
The devices also make better use of WAN bandwidth, driving down the need to expand the capacity of connections among sites.
The client performs the same basic functions as the WANScaler appliances — that is, compressing traffic; optimising the performance of chatty, inefficient protocols; and making TCP respond more rapidly to congestion.
WANScaler client can optimise Common Internet File System, HTTP, network file server and FTP traffic.
The software stores patterns of data sent across the WAN, caching the chunks so they don’t have to be re-sent the next time they come up in a transaction. Customers can set how large they want these caches to be based on available disk space. These dictionaries of data patterns build themselves and when the space is used up, they purge the data that was accessed least.
These clients can work in conjunction with VPN clients, doing their work before the traffic is pushed through the VPN tunnel.
Citrix has also announced WANScaler 4.0 software that adds QoS based on five queues. A new network interface card (NIC) is available for WANScaler appliances that add WAN interfaces so they can home in on two WAN connections.
The software also stores data about traffic that passes over the WAN for a longer time period. Previously it held data from one day, and now it holds a month’s worth.