The world, how it works, surroundings, myself, etc.

Thursday, August 03, 2006

Server Room Correction #2

After sending the long mail on Tuesday night, Paresh and I met Dr Bruhadeshwar on Wed (Aug 2) morning. The meeting lasted long and I must admit that Dr Bruhadeshwar has a lot of positive energy.

The Meeting
I have a feeling that Dr Bruhadeshwar didn't believe much in what I said, at least towards the beginning. This I feel is natural since he has taken over as a CRC head only a month ago and is not acquainted with many things. He had a printout of the list of issues which I had mailed yesterday and it had a lot of markings (even though he mustn't have gotten much time since yesterday night to this morning). He began going through them and discussing corrective action. I had listed only 10 such examples. Even though each of them were tragic blunders in themselves, I mentioned that these were only specific examples of mismanagement and that we needed more fundamental level reforms or else such a list could be practically endless and tackling them one-by-one would be impossible. He did not agree quite, or at least so I feel, towards the beginning of the meeting.

We went through a couple of the specific examples and we landed to the OBH lan problem. That's the time when he called up the server room twice regarding this issue and asked for what was happening. The immediate fallout was that OBH lan was fixed in merely 3 hours (aah.. why couldn't they fix it earlier rather than wait for a bashing for 3 weeks?).

On the issue of poor internet he blamed DC++ for choking up the bandwidth. Now, this is not true. Perhaps he had some misconceptions about the network structure. Eventually, I was able to convince him that DC++ or any such sharing over the LAN doesn't affect internet bandwidth by as much as a dot.

I reiterated that we were just going through specific examples whereas what we should instead be doing is prepare a system. These inconsistencies would just fade away. I gave the example of misconfigured DNS leading to mail-loss by several faculty members. From this point onwards, I believe, he started feeling that I was talking sense. We didn't discuss specific examples after this point.

Student Sysadminship
I have been reiterating that the server room has to have some responsible expert-level administrator, whether student or otherwise. Since such a person is missing at this point, he mentioned that the staff should be trained for it. But this takes time, even if it's possible.

He mentioned that the idea of student sysadmins abandoned because students didn't document properly what they did. Even though I'm not nearly a software engineering fan, I agree that a certain level of book-keeping is required to ensure awareness amonst the staff. However, I categorically mentioned that the level at which it'd be helpful (in the current case) would be an overkill. I don't want to start teaching how to use VIM! I mean, isn't there a documentation/manual available for these?

I do belive that by the end of the meeting, he did start seeing that I had valid points, and also that students sysadmins are required. Probably, this issue will be discussed in some faculty meeting later. Erstwhile, he suggested that there be a knowledge tranfer session (whatever that means) on next Monday where I run the staff through some routine admin know-hows.

I told about the solution which I had proposed, the three steps that needed to be taken (as per my earlier mail). He was convinced.

Personally, I was irritated not with the idea of knowlege transfer but with the term itself -- it's so SE-ic (and need I mention that I don't like SE?). But should I expect that all of a sudden someone who I didn't even recognize (and vice versa) follow me totally? Was Rome built in a day? I don't think so. So, I agreed to the knowlege transfer session to be held next Monday.

As of now, things look extremely positive. Probably, student sysadmins will get appointed. By the end of the meeting, we landed up discussing about the stuff on DC++. Now, I don't watch a lot of movies, but I'd still like to have the option. So, I responded with what a student should and he responded back with what a faculty should -- perfect ethics, although on a light note. Dr Bruhadeshwar is extremely active and I hope things shape up well. Or else, god save the system administrtion, and I'm not just exaggerating.