With all the acronyms floating around in storage discussions these days — and with new ones seemingly popping up on a daily basis — it can be quite difficult keeping on top of them all. As such, we often get emails from readers asking about many of these mysterious acronyms and what they mean to network storage. Sometimes understanding what the acronym stands for is enough to gain some understanding of the technology; other times it doesn’t help much at all. In the next two Storage Basics articles, we’re going to uncover a few of these acronyms, starting with FCIP, iFCP, SoIP, NDMP, and SMI-S.
When we spell out the acronym FCIP, Fibre Channel over IP, we get an idea of what the protocol is designed for. FCIP represents two separate technologies designed to address storage networking requirements as well as the need to network over large distances. The first component is Fibre Channel. Fibre Channel is an established technology optimized for storage-data movement, interoperability, and proven applications for localized storage networking. The second component, Internet Protocol (IP), is a mature technology with a proven ability to transport data over WAN distances.
FCIP combines the best features of both Fibre Channel and the Internet Protocol to connect distributed SANs. FCIP encapsulates Fibre Channel and transports it over a TCP socket. FCIP is considered a tunneling protocol, as it makes a transparent point-to-point connection between geographically separated SANs over IP networks. FCIP relies on TCP/IP services to establish connectivity between remote SANs over LANs, MANs, or WANs. TCP/IP is also responsible for congestion control and management, as well as for data error and data loss recovery.
The advantage of FCIP is that it can use TCP/IP as the transport while keeping Fibre Channel fabric services intact. This allows organizations to leverage their existing technology investments by extending the Fibre Channel fabric over an IP-based link. In many modern network environments, both the IP infrastructure and the expertise are already in place to manage the IP component.
Page 2: iFCP
Often confused with FCIP is the closely named iFCP. The Internet Fibre Channel Protocol (iFCP), however, is an entirely different technology. iFCP allows an organization to extend Fibre Channel storage networks over the Internet using TCP/IP. As with FCIP, TCP is responsible for managing congestion control as well as error detection and recovery services.
The differences between the two technologies are straightforward. FCIP is used to extend an existing Fibre Channel fabric with an IP-based tunnel, allowing networking over distances. This means that the FCIP tunnel is IP-based, but everything else remains Fibre Channel.
iFCP, on the other hand, represents a potential migration strategy from current Fibre Channel SANs to future IP SANs. iFCP gateways can either complement existing Fibre Channel fabrics or replace them altogether. iFCP allows an organization to create an IP SAN fabric that minimizes the Fibre Channel fabric component and maximizes use of the company’s TCP/IP infrastructure.
Storage over IP (SoIP)
Another technology that harnesses IP-based storage is known as Storage over IP (SoIP). SoIP refers to the merging of Fibre Channel technologies with IP-based technology. As mentioned when discussing iFCP and FCIP, merging Fibre Channel technology and IP allows high availability and high performance storage solutions over great distances. SoIP uses standard IP-based protocols, including Open Shortest Path First (OSPF), Simple Mail Transfer Protocol (SMTP), and Routing Information Protocol (RIP).
As you can imagine, using familiar IP-based protocols makes SoIP highly compatible with existing Ethernet infrastructures. For those wondering about how SoIP differs from technologies such as iSCSI, the difference is in the IP transport protocol used. iSCSI uses the TCP protocol for transport, while SoIP uses the User Datagram Protocol (UDP).
TCP is a protocol that provides connection-oriented (guaranteed) delivery of packets across the network. Unlike TCP, UDP offers a best delivery mechanism for packets. As such, it offers lower overhead and therefore more efficient transport. UDP is a connectionless protocol and does not guarantee the delivery of data packets. UDP is used when reliable delivery is not necessary (i.e. when another protocol or service is already responsible for handling this).
Because of the use of UDP, SoIP data transport is faster, yet more unreliable, than iSCSI. The goal of SoIP, like other IP storage options, is to use an existing IP infrastructure to reduce additional hardware costs and retraining.
Page 3: NDMP
The final technologies we will review in this article are the Network Data Management Protocol (NDMP) and the Storage Management Initiative Specification (SMI-S). Today’s network environments are becoming increasingly heterogeneous with multiple hardware and software vendors represented. From time to time, operating systems are upgraded, and over time there is a diverse range of backup media technologies and devices used on a network.
In such environments, backing up and restoring data can become a management nightmare, as each software and hardware backup product can interact with applications in different ways. NDMP is designed to facilitate interoperability in these types of heterogeneous environments. In a typical backup configuration, a backup occurs from the server to a backup device with the backup software controlling and managing the entire process. Individual software vendors use their own protocols to manage the backup data transfer.
In an NDMP backup configuration, the backup data flows through the server to the backup device using a common interface, regardless of the backup devices used or other hardware and software considerations. NDMP is an open network protocol that effectively standardizes the functional interfaces used in the backup and restore process.
NDMP is based on a client/server architecture that is comprised of three separate components: the NDMP host, the NDMP client, and the NDMP server. The NDMP host is the primary device that stores the original data. A NDMP server will then run on the NDMP host and is responsible for managing the NDMP operations. The NDMP client is the backup management software that controls the NDMP server.
To learn more about NDMP and NDMP specifications, refer to the official NDMP Web site at www.ndmp.org.
SMI-S is a relative newcomer as well, but is expected to become a significant component for managing heterogeneous computing environments. Developed by the Storage Networking Industry Association (SNIA), SMI-S is based on the Common Interface Model (CIM) and Web-based enterprise management (WBEM). The primary function of SMI-S is to simplify the administration of complex storage networks by allowing interoperability and integration of hardware and software.
SMI-S provides the ability to manage a heterogeneous storage network from a central location and eliminates the need to manage each device with a separate management application. As an added benefit, the increased interoperability gives organizations the ability to purchase any SMI-S SAN device, regardless of manufacturer, without having to worry about whether or not it will work with other vendors’ products.
For more information on SMI-S, refer to the Web site at http://www.snia.org/smi/about.
In this article we’ve reviewed FCIP, iFCP, NDMP, and SoIP. The next Storage Basics article will continue looking at some of the more promising emerging SAN technologies, including Infiniband, VI, and DAFS.