Editor’s note: This article was initially published in The Daily Gazette, Swarthmore’s online, daily newspaper founded in Fall 1996. As of Fall 2018, the DG has merged with The Phoenix. See the about page to read more about the DG.
System Administrator Angela Andrews came to Swarthmore at 7:00 a.m. Tuesday and immediately noticed something was wrong. The NIC, or network interface card, was in the process of failing. Acting quickly, she brought down the servers and cleanly began a transition to a new redundancy system ITS designed and implemented over the summer.
“Last year,” explained Associate Director Robin Jacobsen, “the system would have been down much longer—even late into the afternoon.” Today, however, by 11:30 ITS was testing the backup and the system was running without a hitch by 1:00 p.m. Hewlett-Packard arrived around 3:00 p.m. with a replacement card.
It appears that the NIC was victim of arbitrary hardware failure. Jacobsen did not expect a repeat of the problem in the near future.
The emergency proved a good test of the new back-up data center tucked away in the basement of Parrish. The system dramatically decreased down-time, exactly fulfilling its duty.
Just out of curiosity- how does one tell if the NIC is in ‘the process of failing’? Is it a visible/tangible thing or something observed via a computer interface?
I’m not an expert, but from my experience with personal computers and NICs, if a piece of physical hardware fails it’s fairly catastrophic. I think you almost totally lose network traffic, and if after you eliminate other potential sources of problems (like the ISP, network conditions or something software related) it’s still unexplainable you just assume it’s hardware and swap it out to see. But maybe there’s something more sophisticated others can comment on. I would like to know what the diagnostic process was, if only for personal edification.