The Metadata Network

As with any client/server protocol, SNFS performance is subject to the limitations of the underlying network. Therefore, it is recommended that you use a dedicated Metadata Network to avoid contention with other network traffic for certain underlying network configurations. The following contains the general rules for the StorNext metadata network:

Note: If the primary MDC is also doing heavy network work, like pushing data to an AWS location, then you should use a DDM if the metadata plus data from the MDC exceeds the bandwidth of the network connection.

Neither TCP offload nor jumbo frames are required. Quantum recommends you have all SNFS clients directly attached to the same network switch as the MDC systems. Any routing required for metadata traffic will incur additional latency that may impact performance. You can use a tool like netperf to help verify the Metadata Network performance characteristics. For example, if netperf -t TCP_RR -H reports less than 4,000 transactions per second capacity, a performance penalty might be incurred. You can also use the netstat tool to identify TCP retransmissions impacting performance. The cvadmin "latency-test" tool is also useful for measuring network latency.