Post reply

Using Triggers

  • June 20, 2005 at 1:25 pm

    #566858

    Sorry, I noticed the Double post and as you can see I deleted it

     

     


    * Noel

  • June 20, 2005 at 1:27 pm

    #566859

    Hey Noel,

    I'm open to any suggestions that anyone has and instead of not liking it....i'm actually very happy and glad that you all have taken the time out to read and respond to the issue i'm having.  So thanks for your suggestion and i am definitely going to take a look at the design (that someone else did)....it's just not priority on the list right now since it would take some time for that and i needed a quick solution. 

    Thanks Again.

  • June 20, 2005 at 1:30 pm

    #566862

    I know the feeling. It is almost a law in software:

    "There is never time to make it right, there is always time to do it over"

    Cheers,

     


    * Noel

  • June 20, 2005 at 1:37 pm

    #566864

    It's hard to tell wether or not the design is wrong... But 50 tables views!! for a single query seems like a lot to me.

    Is there anyway that you could change the program so that less information is shown at the same time???

    Like pick a group of project, select a project(s), then show the relevant information separatly (like cost, clients, workers..)????

    The most complex report I did here only had 5 subreports. Which is something like 12 tables, 5 Stored Procs and bam you're done (the most complexe by so far).

  • June 20, 2005 at 1:41 pm

    #566868

    So true... wasn't that quote from Boston Public or another show?

  • June 20, 2005 at 1:43 pm

    #566870

    That's the problem....this area is for upper level managers of the company and they want to see a lot of different data fields without having to click here and there.  So it's a one click approach where they get to see all the data that they want for all the different types of projects (grouped into one). 

  • June 20, 2005 at 1:44 pm

    #566871

    It can be possible to have that many tables (I've dealt with 1000's), I had to deal with the 240 limit many times because of the views of views of views... issue

    95% of the time a rewrite of procedures and change on table structures was a more stable solution than pacthing the already patched code

    just my $0.02

     


    * Noel

  • June 20, 2005 at 1:46 pm

    #566873

    Ouch... is it possible to see such a query?? I've never seen one that big .

  • June 20, 2005 at 1:48 pm

    #566875

    They preffer scrolling through 1000s of rows and a few dozen columns to find data instead of doing 2-3 simple clicks and seeing only what they need????

    Are you sure you have the right managers for the job?????????

  • June 20, 2005 at 1:51 pm

    #566876

    Unfortunately, I can't do anything ( posting realated ), I signed a confidentiality agreement and not even the name of the company can be said here ...  they can still provide me some business

    The reason also is that sometimes part of the query was automatically generated and you will get those 50-80 join monsters

     


    * Noel

  • June 20, 2005 at 1:54 pm

    #566878

    I see. How fast was that monster running anyways??

  • June 20, 2005 at 2:26 pm

    #566892

    To be honest those monsters are runned at scheduled times(because Everybody would suffer it if they were ad hoc) and hopefully nobody happen to invoke them at unexpected times. It has happened and only one is needed to bring the server to its knees for a while

     

     


    * Noel

  • June 20, 2005 at 2:29 pm

    #566895

    How long does it take to run one of these?

  • June 20, 2005 at 3:09 pm

    #566902

    It was very marked the difference between picked activity and non pick hours:

    When in normal(working hours) could drag the system for about 30min

    When scheduled (at night) 5 min will suffice

     


    * Noel

  • June 20, 2005 at 8:29 pm

    #566924

    Hehe, I guess that 5000 less query/hour will do that to a server .

Viewing 15 posts - 16 through 30 (of 31 total)

You must be logged in to reply to this topic. Login to reply