Results 1 to 8 of 8
  1. #1
    Join Date
    Sep 2003
    Location
    Texas
    Posts
    9

    Unanswered: netrwork database load time

    scenario:
    20mb shared mdb-peer peer ntwk-6 users

    The loading time increases as more users have db open (no activity..just open)

    for ex:

    if no sta has the db open, the open time is very short for the first sta to open db.

    if 5 stations have db open (no activity...just open), the 6th sta experiences a 13 sec opening time

    the opening time gets worse as more stations have the db open!!

    Can someone explain this??

    Thanks...bed78552@yahoo.com

  2. #2
    Join Date
    Nov 2003
    Location
    Netherlands
    Posts
    85
    Has your database a separate front - and back end?

  3. #3
    Join Date
    Sep 2003
    Location
    Texas
    Posts
    9

    network database load time

    no, it's not split...

    Trials that I've run show there,s to gain in loading time by splitting...

    Actually there are two db about the same size...and some table links between them..

    The slow loading time problem is the same with either...

  4. #4
    Join Date
    Sep 2003
    Location
    MI
    Posts
    3,713
    Perhaps it has to do with the internal Access overhead of having to check for concurrency ... Say, checking that the Xth user does not overflow the 255 concurrent user limit ... And other things like that (opening the internal ADO and DAO connections comes to mind also - CurrentProject and Currentdb) ...
    Back to Access ... ADO is not the way to go for speed ...

  5. #5
    Join Date
    Sep 2003
    Location
    Texas
    Posts
    9

    network database loading time

    Strictly DAO myself ...

    I thought that possibly each of the loaded copies of the db must continuously update all other copies and that meant that network traffic increases a lot with each additional copy loaded by a station..

    however, I can't seem to find any clear detailed data as to just how multiple station's copies of db affects the network traffic...

    Anyway, thanks for the input

  6. #6
    Join Date
    Dec 2002
    Location
    Préverenges, Switzerland
    Posts
    3,740
    go DAO!

    i worry about the peer-to-peer stuff: i've never tried p2p with access (to be honest, i gave up with p2p in the first 5 minutes, never mind burdening it with an access problem), but somewhere in the p2p setup, one (at least, i hope only one) of the machines has a.LDB file generated when the first user signs up to your .MDx
    ...it could be interesting to monitor the traffic on that .LDB as you add users.

    i do still have a few shared .MDx backends with three or four users running on a clean net & generous fileserver (NT4 no less!): no observable difference in startup between first or fourth user

    izy
    currently using SS 2008R2

  7. #7
    Join Date
    Sep 2003
    Location
    MI
    Posts
    3,713
    Quote Originally Posted by bed78552
    Strictly DAO myself ...

    I thought that possibly each of the loaded copies of the db must continuously update all other copies and that meant that network traffic increases a lot with each additional copy loaded by a station..

    however, I can't seem to find any clear detailed data as to just how multiple station's copies of db affects the network traffic...

    Anyway, thanks for the input
    No copies ... As izy pointed out, there is a shared lock file that's updated with each user ... Also, if the network db has links, access is doing that over the network also ... Establishing, verifying, and maintaining ...
    Back to Access ... ADO is not the way to go for speed ...

  8. #8
    Join Date
    Sep 2003
    Location
    Texas
    Posts
    9

    Network loading time

    izy...Thanks...I (dummy me) have discovered at least one cause of delay...was using the open event of startup form to do some work...will continue to experiment with it...

    on a side question, I have difficulty seeing any difference in the p2p vs client/server with Access....One of the stations will be the file server for the DB... even if it's split, some machine must serve as the fille server for the backend.....The only difference I can see is that the with p2p, the serving machine might have someone doing work on it, while a 'server' probably would not. Am I not ubderstanding something...

    Woulld like to find some 'plainly written info' describing just how access works on a network in a shared scenario

    I'm foolin around with maintaining performance as the db grows..

    Anyway thanks to all for your input...

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •