Skip to main content

View Diary: A Software Engineer's take on (145 comments)

Comment Preferences

  •  Bandaids are normal (3+ / 0-)
    Recommended by:
    Oh Mary Oh, Kristjan Wager, nargel

    Production down is always fixed like this:

    1.  Get the system up and running
    2.  Diagnose what went wrong (sometimes needed for #1)
    3.  Put in a temporary fix or work-around with the emphasis on SAFE.  Usually the minimum possible to avoid the issue.
    4.  Code a proper fix in with all other bugs/features discovered recently into a build, with proper project management, test cycles etc to ensure the proper fix won't break anything else.
    5.  Deploy that fix along with numbers of other fixes, replacing the bandaids.

    Only if all of your IT is spending all of its time on work-arounds does the scenario you describe occur.  That's a trap, and it happens sometimes with understaffed IT, but it is fairly well understood in most IT organizations these days that each bandage/work-around makes a system more fragile.   I've spoken with one person involved in the back end a little bit offline and he was talking about "Phase 2" two weeks in.  That's IT code for "all the stuff we triaged out of the first build, plus the first round of proper bug fixes".

    It could happen the way you describe, but I judge it unlikely because they've got people working on proper phased rollouts.

    •  My wife is an IT project manager. (0+ / 0-)

      And her previous job was in a place where they WERE spending most of their time on work-arounds because the company bosses hated to let them spend the time to fix it properly and demanded more content to sell. They were a media company.

      If the bosses are really as bad as depicted here, then maybe that will happen.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site