SCSI—Small Computer System Interface

Successfully installing a SCSI device on a Linux PC.

The SCSI (small computer system interface) bus is a system commonly found on workstation-class machines, but is available for other systems such as personal computers (PCs). The advantage of SCSI is that once the controller is set up, you can just add new devices (typically up to seven) with a minimum of fuss. The new device is merely plugged into the existing “daisy chain” bus. The user must ensure that the last device in the chain is properly terminated. We will discuss termination in a later section.

The idea to write about SCSI came out of a project I was working on to provide Linux-based web and network servers for a local materials science company, Fibics, Inc. (http://www.fibics.com/). The network server was to include hardware for performing file-system backups. I am not a SCSI expert and had to learn the hard way, so I thought I'd share the basics in hopes that others might benefit from my experience.

Where Do I Start?

You will need a computer (obviously), a SCSI controller card to match the bus type of your computer and one or more SCSI devices (disk drive, tape drive, etc). Note that different SCSI devices may require different SCSI bus types (SCSI-1 vs. SCSI-2 vs. ultrawide, etc.) and this may affect your choice of controller card. For more information on the different SCSI buses, I recommend the booklet Basics of SCSI from Ancot Corporation (http://www.ancot.com/).

Before purchasing hardware, you should review the relevant HOWTO documents (SCSI, hardware) and the hardware compatibility list for your Linux distribution. The HOWTO documents I read are from the Red Hat distribution and are installed in /usr/doc: /usr/doc/HOWTO/SCSI-Programming-HOWTO and /usr/doc/HOWTO/unmaintained/SCSI-HOWTO. The HOWTO documents provide some useful information, although they are somewhat dated.

If your PC has no SCSI devices or controllers, you should start by determining what device you need. For my project, a high-capacity 4mm DAT drive was required. A review of available drives suggested a Seagate Scorpion (12/24GB) would meet the requirements. Once the device is selected, you can narrow down your choice of controller cards based on the type of SCSI bus the device requires.

The following steps were used to determine what controller card I would require:

  • The target machine was an AMD K6-2-based PC with PCI and ISA buses.

  • A 4mm DAT drive was needed for backup purposes.

  • The selected drive required a SCSI-2 bus.

  • An Adaptec 2910 controller supports SCSI-2 and is a PCI card; it is not specifically listed as incompatible in the hardware list.

  • A driver for the 2910 was needed to run under Red Hat 5.2.

The most difficult task is determining the correct driver, as such information was not documented anywhere I looked. Seagate has an automated phone system (1-800-SEAGATE) which provides considerable information about their hardware. Unfortunately, on the software side only Windows 95/98/NT information was available.

The solution was to post a message to the local Linux mailing list. Many thanks to the individuals on the Ottawa Carleton Linux Users Group (OCLUG) mailing list for indicating that the “aic7xxx” SCSI driver module was the proper choice. The OCLUG URL is listed in Resources.

Another post to the OCLUG mailing list and some looking around led me to more detailed information. README files for the various SCSI drivers reside in /usr/src/linux-2.0.36/drivers/scsi/. Here, you can determine what cards or chip sets are supported by each driver. It would be worth looking at these files prior to purchasing a SCSI controller card.

To add the required driver, I started up Red Hat's control panel. I selected “Kernel Configurator” and clicked on “add”. A “Module Type” window opened; there I selected “scsi_hostadapter” from the “Module Types” pulldown menu. I clicked “OK” to open the “Module Definition” window, then clicked the “Which module” pulldown and selected the required driver. Finally, I clicked “OK” and then “Restart kerneld” in the “Kernel Configurator” window.

Hardware

Having determined what was needed, the hardware was purchased and installed. The Seagate DAT manual was not as clear as it could have been about installing a single SCSI device, but all the required information is there. Listening to the information on their automated telephone help line answered my remaining hardware questions. I'm providing sufficient information here (I hope), so you won't have to make the phone call.

Two rules must be followed when dealing with a SCSI system. One, every SCSI device must be set to a unique SCSI ID number. ID #7 is normally used for the SCSI controller card and ID #0 is often used by the boot drive. Two, a SCSI bus (cable) must be terminated at both ends in this manner:

  • Internal devices use active termination on the last device in the chain, at the end of the ribbon cable.

  • External devices require a terminator plugged into the unused connector on the last device in the chain.

  • The controller card provides automatic active termination. If the card is used only for internal devices, the external connector need not be terminated.

In my particular case, I wanted to install a single internal device. The 2910 controller card was plugged into a free PCI slot on the motherboard. The DAT drive was configured for active termination via the appropriate DIP switch. Other DIP switches were used to set the drive to SCSI address #2. Then the drive was installed in a free 5-1/4-inch drive bay. Since the DAT drive was the only SCSI device being installed, it was plugged into one end of the SCSI ribbon cable. The other end was plugged into the SCSI controller card. The Adaptec 2910 “kit” came with the ribbon cable that allows two devices to be connected to the card. The card also has an external SCSI connector which was not used (nor terminated).

______________________

White Paper
Linux Management with Red Hat Satellite: Measuring Business Impact and ROI

Linux has become a key foundation for supporting today's rapidly growing IT environments. Linux is being used to deploy business applications and databases, trading on its reputation as a low-cost operating environment. For many IT organizations, Linux is a mainstay for deploying Web servers and has evolved from handling basic file, print, and utility workloads to running mission-critical applications and databases, physically, virtually, and in the cloud. As Linux grows in importance in terms of value to the business, managing Linux environments to high standards of service quality — availability, security, and performance — becomes an essential requirement for business success.

Learn More

Sponsored by Red Hat

White Paper
Private PaaS for the Agile Enterprise

If you already use virtualized infrastructure, you are well on your way to leveraging the power of the cloud. Virtualization offers the promise of limitless resources, but how do you manage that scalability when your DevOps team doesn’t scale? In today’s hypercompetitive markets, fast results can make a difference between leading the pack vs. obsolescence. Organizations need more benefits from cloud computing than just raw resources. They need agility, flexibility, convenience, ROI, and control.

Stackato private Platform-as-a-Service technology from ActiveState extends your private cloud infrastructure by creating a private PaaS to provide on-demand availability, flexibility, control, and ultimately, faster time-to-market for your enterprise.

Learn More

Sponsored by ActiveState