"You can't manage what you can't see" is a popular axiom in the network management industry. The idea is that the foundation of any organization's management strategy should be to maximize visibility across the network. Only by "seeing" traffic flows can IT departments truly understand the IP network and troubleshoot problems quickly. This is one of the reasons visibility vendors like Gigamon have been red hot of late.
But what about the Fibre Channel storage area network (SAN)? Despite the rise in IP network visibility products, there hasn't been any innovation in helping organizations understand what's happening on the SAN. So if it's true that "you can't manage what you can't see," then the lack of SAN visibility means that there's really no way to manage the SAN.
This morning, Brocade, the market leader in SAN network infrastructure, released a product it's calling the Brocade Analytics Monitoring Platform that is a fully integrated monitoring and analytics solution for Brocade Gen 5 (16 Gig) Fibre Channel fabrics. The product is a 2U, 24-port management, purpose-built appliance that runs Brocade's Fabric OS (FOS) to ensure tight interoperability with the SAN.
The Brocade product sits out of band and connects into a single port on any Brocade Gen 5 FC director or switch. The out-of-band connectivity enables the product to collect data from storage or host ports without impacting performance levels or compromising security. Through that port, the product can capture and analyze millions of traffic flows and give detailed performance metrics for storage and server connections within the SAN fabric. The metrics captured are a mix of real-time and historical and include the following:
- Read/write I/O latency and first response times (including both average and maximum values)
- Read/write IOPS transfer rate
- Other command latency stats (reserve, release)
- I/O queue depth (pending IOs)
- Protocol errorstats: Check conditions; I/O aborts; SCSI reserves, releases, reserve conflicts
- Fabric latency
Unlike most visibility tools that require a third-party application to view the data, the information captured by the Analytics Monitoring Platform can be viewed in the Brocade Network Advisor software platform and can generate reports to show granular reports, as well as trending information based on up to two years of historical data stored in Networks Advisor.
For IT departments, the capturing and analyzing of data is key to understanding performance and shifting to a predictive management strategy. Once the data is collected and analyzed, baselines can be set and then watched. If certain flows are trending in a direction that will eventually impact performance, actions can be taken to remediate the issues before they become problems. Many IT departments I talked to want to be more proactive with their management strategies, but pro-activeness cannot be achieved without the tools to predict when problems occur. Without the proper baselines and trending information, the best a business can hope to achieve is an educated guess. This may be OK if your lead engineer is Mr. Spock, but without the calculator-like brain of a Vulcan, you're just shooting from the hip.
Sign up for CIO Asia eNewsletters.