With the advent of 128-, 140- and 256- port count Fibre Channel switches and switching directors, previously complex fabric topologies have been simplified. Even so, there is the need in some environments to scale beyond the limits of a single device or fabric.
If you're currently reviewing your SAN topology with an eye toward scaling beyond a single fabric, there are a variety of technologies and products that can help you accomplish your goal. There are also a few pitfalls to watch out for, both during the initial assessment and the actual deployment. Here are a few things to keep in mind:
I. Reasons for scaling beyond a single fabric
II. Possible architectures
III. key technologies
Even if you are not planning on combining devices from different fabrics, some upfront planning now can reduce, if not eliminate, future interoperability challenges. Remember that:
- Routing is not just a technology for spanning distances. It can be equally of use in a local environment, for example, segmentation of physically interconnected SAN islands and protocol conversion to support iSCSI-enabled servers.
- Plan and design for growth, maintenance and diagnostic of your storage network.
- Utilize the applicable scaling technology and approach to meet your specific needs.
- Use SRM tools to keep track of and monitor the health and status of your storage network, including performance and resource usage. Change control and management software can help to proactively determine and isolate problems before they occur.
- Utilize unique, descriptive naming and addressing conventions including domain IDs. This is helpful regardless of if you are merging fabrics or keeping them isolated.
- Enable domain ID lockdown to insure unique domain IDs on switches, particularly for FICON cascade, that also require fabric binding security.
- Isolate traffic and workload with segmentation, zoning, LUN masking and mapping.
- The number of ISLs between switches is a function of how much bandwidth is needed for your applications and latency concerns.
- Beware of lost ports and overhead to support horizontal scaling and expansion.
- Configure additional ISLs for load-balancing and redundancy. The level of performance you need in a failure situation (loss of an ISL, loss of a switch, or loss of a communication circuit) will determine how to spread your ISLs across the fabrics.
- Techniques that can be used to increase bandwidth between switches in a fabric, and between fabrics, include ISL aggregation (trunking), physical multiplexing using dense wave division multiplexing (DWDM), ISL grouping with load balancing and faster ISL interfaces (4 Gbps, 10 Gbps).
Checklist: Five steps to maximizing disk capacity
Storage Boot Camp: One-stop training for storage pros
About the author: Greg Schulz is a senior analyst with the independent storage analysis firm, The Evaluator Group Inc. Greg has 25 years of IT experience as a consultant, end user, storage and storage networking vendor, and industry analyst. Greg has worked with Unix, Windows, IBM Mainframe, OpenVMS and other hardware/software environments. In addition to being an analyst, Greg is also the author and illustrator of Resilient Storage Networks, Greg has contributed material to Storage Magazine. Greg holds both a computer science and software engineering degree from the University of St. Thomas.