On November 2, Seton Hill University (SHU) was offline from approximately 2:30 a.m. to about 9:30 p.m. The following Saturday, the Internet was down again in the afternoon.
“If the Internet is down for 19 hours, that’s practically the whole day. I really don’t think it should take a whole day to fix the network,” said Amanda Horner, a junior.
Some students tried earlier in the day, found the Internet in their dorms not working, and crammed into the computer labs in Maura since the Internet came back to the computer labs before the dorms. Some even called campus police when their Internet was gone. And at 10:09 p.m., a global email from Chris Sherman, of the IT department, identified the cause as the “…temporary failure of one of our hobnob appliances.”
By Everett Manns
Contributor
On November 2, Seton Hill University (SHU) was offline from approximately 2:30 a.m. to about 9:30 p.m. The following Saturday, the Internet was down again in the afternoon.
“If the Internet is down for 19 hours, that’s practically the whole day. I really don’t think it should take a whole day to fix the network,” said Amanda Horner, a junior.
Some students tried earlier in the day, found the Internet in their dorms not working, and crammed into the computer labs in Maura since the Internet came back to the computer labs before the dorms. Some even called campus police when their Internet was gone. And at 10:09 p.m., a global email from Chris Sherman, of the IT department, identified the cause as the “…temporary failure of one of our hobnob appliances.”
“It’s very frustrating,” said Christine Rauch, a sophomore who needed to find a manuscript for class.
Some students did not take particular issue with the Internet being down and accept the general level of reliability in our school’s systems. Dawna Gilvarry, a freshman, said, “Part of the reason I chose to come to Seton Hill was that it wasn’t overly computerized.”
According to Sherman, the repeated poor performance of the network was just a fluke and there is no specific reason to believe a pattern of network failure is in the future. His automated system monitor reports that the network actually has had a 99.5 percent runtime since September. Weekends are more difficult because the staff is reduced to a level of “on call” and leave most functions to automation.
On November 2, all the automated systems still showed that the network was functioning, and the IT department didn’t know there was a problem until they got the call.
Eventually they had to manually troubleshoot until they discovered the hard drives of one hobnob, which is part of the server hardware, had been completely filled with “network event logs.” Sherman said that it takes a long time to happen, so it’s unlikely to reoccur this semester.
To stress the effectiveness of the auto self-diagnosis, Sherman wasn’t even aware that the network had gone down until Saturday the next week, and concluded that it basically fixed itself.