Adding a device or changing the length of a SCSI cable can cause a major performance drop in a SCSI system. The...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
obvious culprits are cable and termination problems, but sometimes the problem persists even when those are fine.
Another possible source of problems in a high-speed SCSI bus is reflections. Paralan Corp., a maker of SCSI devices, points out that the signals -- at SCSI speeds -- can bounce back along the cable when they encounter the next device in the change. In this scenario, the SCSI cable acts as a transmission line and can be subject to standing waves and reflections in the electrical signals that travel down the cable.
Reflections can be caused by a number of factors, including improper terminations and poor backplane design, Paralan says. A less obvious culprit is the length of the SCSI cable itself.
If the devices are the correct distance apart -- and the signal is strong enough -- the device that puts data on the bus may interpret the reflection as an incoming signal. Since the reflection isn't the signal the transmitting device expects, that device may resend the data, or even shut down the bus.
Untangling all this can slow transmission speeds over the bus considerably. It also causes performance to suffer.
If you're not expecting it, reflection can be a particularly difficult problem to troubleshoot because the connections, terminations, and other parts of the SCSI system come through tests with perfect marks.
To address a potential reflection problem, try changing the length of the SCSI cable, Paralan says. This changes the timing on the reflected signal and makes it easier for the device to ignore it.
For more information, check out Paralan's SCSI FAQ, at http://scsifaq.paralan.com/. Or, if you're new to the storage field, look up the terms "SCSI", "bus", and "backplane" on our Whatis.com sister site http://www.whatis.com/.
Rick Cook has been writing about mass storage since the days when the term meant an 80K floppy disk. The computers he learned on used ferrite cores and magnetic drums. For the last twenty years he has been a freelance writer specializing in storage and other computer issues.