What is the expert---can solve the problem, no matter which link has the problem

English Original: The best debugging Story I ' ve ever Heard

That was the beginning of the 80 's, when my father worked for a storage device company, which no longer exists, it produces tape drives and a pneumatic system that drives these tapes at high speeds-a product of that era.

(used under license from Laughing Squid. The original image can be found here. )

They retrofit the tape drive so that you can have only one central drive-the "A" disk-connected to several "B" disks, which resides in a small operating system in the memory attached to a disk and is responsible for the data read and write operations of all B drives.

Each time you start a drive, you need to insert a floppy disk into your peripheral drive, and the operating system will load a disk into memory. The operating system is surprisingly simple-its processing power is generated from a 8-byte microcontroller.

The target users of such devices are companies with large amounts of data-banks, magazines, and so on-that they need to print large numbers of address books or bank accounts.

A customer has a problem. During printing, an individual drive stops working, causing the entire printing process to terminate. In order to reload the drive, the duty officer must restart all drivers-if this happens in a 6-hour print job, a lot of valuable computer usage time will be wasted and the entire task will not be completed by time.

The company dispatched a technician. The technician did his best and couldn't reproduce the problem in the test environment: the problem appears to be only in the process of printing a lot of tasks. Although the problem was very unlikely on hardware, he replaced all the devices--memory, microprocessor, disk drive, all the parts associated with the tape drive--but the problem still arose.

So the technician called the headquarters to call an expert.

The expert asked for a chair and a cup of coffee, and sat in the computer room--at the time they had a computer room dedicated to the computers--a lot of print jobs were prepared by the watch, and he was watching. He waited until the machine collapsed. The machine really collapsed, and everyone looked at the experts-the experts didn't find any clues. He ordered the print task to be re-executed once, and all the on-duty personnel and technicians returned to their respective positions.

The expert was again on the chair, waiting for the machine to crash. This is six hours, but it really happened again. Experts still haven't figured out what caused the crash-except for one thing he noticed was that the crash always happened when the house was a lot more. He ordered to print again, to sit back and wait.

When the third crash, he found something. The crash always occurs when the on-duty person changes other boot disks that are not associated. Further, he realized that a crash would occur when a person on duty walked across a floor.

The floor is made of aluminum plate, below the 6 to 8 inches high of the compartment, the computer used a large number of cables are under the floor, so you can avoid the duty personnel accidentally kicked them. The ground plate is very close together, this is to ensure that the rubbish does not fall into the space of cable passage.

Experts say a floor has been deformed. When the duty officer stepped on the corner of the deformed floor, the edges of the plate rubbed against each other, which generated static electricity between the plastic that connected the plates, and caused electromagnetic interference.

Today, all RAM has anti-EMI capabilities. But there was no such technology at the time. Experts point out that electromagnetic interference destroys the work of RAM, and the operating system crashes.

The expert called the maintenance department, brought a new floor, he put it on his own, and the problem was solved.

http://kb.cnblogs.com/page/514193/

What is the expert---can solve the problem, no matter which link has the problem