Bridgeworks is a UK based storage protocol network bridging specialist. Recently it developed bridges for Fibre channel, iSCSI and SCSI. The company recently briefed me about eliminating a significant amount of elapse time at customer CVS in the US.
TCP/IP latency
The product range is called SANSlide and is speeding up data transmission over distances. Latency is conventionally viewed as a given but engineers found a way to chip away at the inner workings of network transmission. Which is seen in the case of WAN optimisation by companies such as Riverbed or Replify my previous blog post where chattiness and added data could be eradicated. There is another in-efficiency in TCP/IP because the transmission eventually endures a wait before more data is attempted to be transferred.
SANSlide
Bridgeworks SANSlide takes that transfer stream and breaks it up across several virtual channels before it is sent onto the WAN. Another SANslide takes the channels and puts them together into the original stream again at the other end. The virtual channels are being managed by an AI engine on the SANSlide. There currently versions for Fibre Channel, iSCSI, SAS and SCSI. They are marketed as solutions for long distance storage transfers of large data amounts for backup/restore and DR/BC.
Benefits
In the customer case story CVS achieved a 50 time reduction in its time of almost 15 hours to backup 50GB. Customers may already have implemented de-dupe based solutions but the TCP/IP issue still remains. There are entry level, midrange or enterprise versions of SANSlide and they are competitive compared to high calibre network alternatives.
