Browse Prior Art Database

Storage Networking Problem Determination

IP.com Disclosure Number: IPCOM000022539D
Original Publication Date: 2004-Mar-20
Included in the Prior Art Database: 2004-Mar-20
Document File: 2 page(s) / 41K

Publishing Venue

IBM

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Storage Networking Problem Determination

The price, performance, and scalability characteristics of the Fibre Channel SAN
apply across all information technology platforms, making it an interconnect
applicable to multi-vendor desktops, servers/clusters, mainframes and super
computers. Through protocol encapsulation, a single fiber can simultaneously
support existing I/O channels (eg., SCSI, IPI, HiPPI, FICON, and Block Mux),
networks (TCP/IP), and memory-to-memory protocols (VI), providing a
universal
interconnect
environment while also preserving investments in existing system
hardware and software. This aspect of the FC SAN has driven the concept of
"shared heterogeneous storage consolidation" in which multi-vendor storage
elements and hosts share a common SAN fabric. This is the great promise of FC
SAN - large scale heterogeneous, inter operable, networked hosts and shared
storage elements, able to be managed from a single point.

Along with this promise comes the unexpected high cost of SAN link and component
failures (link failures are primarily due to optical components) and the
associated cost of serviceability (the S in RAS) which is a

byproduct of the complexity and the heterogeneity of the SAN. The heterogeneous
or "open" nature of the FC SAN has helped foster the view that there is no
single owner for SAN serviceability, ie., each product vendor supplies RAS
capabilities on a component basis, and that RAS and more importantly,
serviceability does not exceed the scope of the local component. But just as the
need for SAN management and security took a number of years to be realized by
the SAN industry, the realization that SAN serviceability is a fundamental
requirement is also late in coming. The bottom line is that the SAN industry has
ignored the need for a SAN Serviceability Architecture to this day. When this
shortcoming is coupled with large complex SANs comprising thousands of links,
the ability to quickly service SAN link faults and thus guarantee system
ava...