Monday, June 20, 2005

Timeline - My Weekend

The past 2 days have been sooo damn hectic. My guide asked me on Friday evening to modify some of my earlier course project code to work with some BerkeleyDB input, and said it should take me a few hours to modify, and a few hours to run the code (expected running time=4hours).

9 p.m. Friday : Slept in the lab on a chair waiting for details about the modification from my guide.

11.30 p.m. Friday : I've got mail !!!

11.35 p.m. Friday : Start looking at code sent, documentation, my old code...

2 a.m. Saturday : Modifications begin - I realize we have some data format issues, add a new hash table, then realize this has introduced more unexpected issues - undo previous work :(.

8 a.m. Saturday : Go to hostel - breakfast & bath - no sleep in bed yesterday night.

9 a.m. Saturday : Continue work - 4 files modified and successfully tested :) - but 2 big ones to go.

11 a.m. Saturday : Status meeting with guide. Issues resolved - deadline postponed.

4 p.m. Saturday : I added 1 more hash table , removed dependancy of code on lucene - skipped tiffin & went late for dinner.

3 a.m. Sunday : It runs !!! But tested only on small input. Start running on required big data set.

3.30 a.m. Sunday : Fall asleep in lab on chair, while program is being executed.

6 a.m. Sunday : Get up to realize program crashed (God knows why) - back to hostel for breakfast, & bath - have not seen my hostel bed for quite some time now.

8 a.m. Sunday : Start running the code again to confirm it crashes :).... try debugging but am tooo tired to think. My project partner is back from his Raigad trek, starts suggesting possible failure reasons.

11 a.m. Sunday : Meeting with guide, followed by crazy debugging. Program crashes after 30 CPU minutes, so we sleep in 30 minute spells while waiting for it to crash during debugging.

7 p.m. Sunday : Realize fault is of outdated BerkeleyDB version - download new version, start running it.

9 p.m. Sunday : Hostel for bath, change of clothes - come back to see program still running - bug gone :))).... 4 more hours of running time left, so go out for dinner.

12 a.m. Monday : Program still running, go to sleep. Its not getting enough CPU time due to heavy duty programs being run by my guide and Sujatha.

5 a.m. Monday : Expect to see it completed, but unfortunately find out-of-memory exception instead :((. Considering that our system is a 2 CPU, 4GB RAM machine, that comes as a huge shock (HOW???? :O)

5.30 a.m. Monday : back in lab, fixing code to ensure earlier 5 hours of processing does not go to waste. Restart program from mid-point and hope that consistency is maintained.

8 a.m. Monday : breakfast, blogs, start writing this post.... current status : it should finish processing in another 5 minutes - then we confirm that output consistency is maintained....

10 a.m. Monday : It runs & it is consistent atleast in all our tests so far :)). Have mailed guide. Now trying to find performance stats.... hopefully, this will be the end, and afternoon will be a movie and sleep :)).

Man, what a weekend.

P.S. : Just in case you are shocked by my dedication & hard-work, let me tell u, I am dedicated and hard-working and all that in bursts of 3-4 days every 6 months or so...
By afternoon, I should be back to procrastination :)).

Another P.S. : Sabah came to IIT yesterday but I was too busy to show her anything except our lab, or spend much time talking to her. :( Felt v guilty - sorry Sabah - will show you around IIT next time. :)

3 comments:

  1. Some schedule.4gb memory? all day work??
    Reminds me of Digital Fortress..but you arent Susan Fletcher are you?? :-P

    ReplyDelete
  2. hey atleast u cracked it... nice to hear... and please do not feel guilty... i had fun anyways...

    ReplyDelete
  3. im sure she would love to be susan fletcher ......... and hey did u know my nick name is david becker.....

    zeus

    ReplyDelete