Troubleshooting Technology Problems

Technology is not perfect and sooner or later when working with technology one will face obstacles that must be overcome. I think I have noticed something about the approach people take in attempting to solve such problems. When working in a new or unmastered area, people first assume that they must be doing something wrong when they encounter problems. Sometimes this lack of creative thinking makes solving the problem impossible.

I encountered one such problem today. I operate a wiki on one of my servers for students in my summer graduate class. My tech skills are developed out of necessity and I always feel like I am operating right at the edge of my capabilities. Not a very confident feeling.

An interesting theoretical position taken by some who design wiki software (in my opinion) is that the forces of good will always overcome the forces of evil. In other words, one can allow all comers to post and edit each others posts and the final product will improve rather than disintegrate. I am not as confident and so I want to impose some access limits on the wiki I operate for my students. The system that mediawiki (the software I use) has available for addressing such issues combines login and email authentication. As system administrator, I can decide to allow a person access to the wiki and I provide that person access by using the wiki to send that person an email with a randomly generated password. If the person is who he/she claims to be (the email address belongs to the person), the person will receive the email and then use the password provided to login to the wiki (and then create a personalized password). Anyway, the system is about providing a reasonable amount of security.

This system was working until last night. A student emailed me and said she forgot her password and needed to create a new account (passwords are encrypted so creating a new account is one solution). I connected to the wiki program as administrator and had the system send her a password. She contacted me today and said the email never arrived. My first reaction was that she must have made a mistake (the wiki said it send the email). After trying a couple of more times without any luck, I began to try my standard debug strategies for such situations. I attempted to enroll myself as a student user and have the system send an email to an account I know works – my own. Still nothing. This is the point at which doubt takes over. Some mysterious config or preference file in Apache, MySQL, or MediaWiki (or some combination) must have gone bad. I had no idea where to begin looking or what I would do about the wiki requirement I had established for me course. I was feeling very helpless.

Then I remembered Port 25. What are the odds a 50+ year old educational psychologists would know what ports are and what port 25 is used for? A better question would be should I have to know what ports are so that I can run a wiki?

I called the computer help desk. “Is the university blocking port 25?”, I asked. “Yes we are” was the reply. From this point on I will have to approximate my comments a bit. “It would have been nice if you would have taken the time to contact those of us who run servers before you did that” I suggested. “I spent a lot of time assuming something had gone wrong with my server.” “We did not know how to find all the servers that are out there” was the justification for noting letting me know that my server would stop sending email without notice.

Port 25 is used for email (80 is for http, 21,22 for ftp, etc.). The university decided that by blocking port 25 those servers responsible for unknown email spam could be eliminated. Sure enough blocking port 25 eliminates email spam should my server have been sending spam. Blocking port 25 also prevented students from gaining access to the wiki to complete course assignments.

Since I am still in a problem solving mode, how would I locate servers so that those individuals responsible might be contacted? I would probably scan for machines on the net with port 80 open, I would use the server names to identify the department associated with the IP address hosting the server (grabe.psych.und.nodak.edu is as one might guess in the Psych department), and I would call or email the chairs of such departments and ask them to inform anyone in the department running a server that the email port would be blocked.

I am feeling better now. Thanks for reading. Remember, if something goes wrong, it may not be your fault. 😉

Loading

Leave a Reply