RobotStudio event

Problems with Large .ldf File [WebWare Server]

Options
kevin.behnisch
kevin.behnisch
in WebWare

image

When performing archiving the database the .ldf-File grows up more and more and the computer slows down. Attached the ldf-File with 32GB size!!!!!. In some sql-forums I read about shrink the ldf file. How could we prevent this behavior in the MS MSDE?  

This behaviour is seen at two customers using custom databases

System Using:

WW Server 3.2 Sp2, Windows 2000 Server, MSDE-Engine

big_ldffile.zip

Comments

  • RussD
    RussD ✭
    Options

    Hello Kevin,

    How frequently is the database being archived?

    How many devices are connected to this WebWare Server?

    How long has this database been in service?

    Russell Drown
  • kevin.behnisch
    kevin.behnisch
    Options

    Hello Russ, we got two customers with these problems

    How frequently is the database being archived? Every four weeks

    How many devices are connected to this WebWare Server? 20 robots (benteler), 2 robots+ 1OPC (ziehl-abeg)

    How long has this database been in service? since February (benteler), since january (ziehl-abegg)

     

  • RussD
    RussD ✭
    Options

    Hello Kevin,

    I would recommend that you do the following:

    1. Detach the current WebWare database as described in the documentation, then move the db to another SQL Server instance or delete it if you don't need the data it contains.

    2. Create  a new, empty database as described in the documentation.

    3. Change the archive schedule to perform the archive much more frequently, i.e. once per week

    4. Examine the database contents to see if it is being filled with useless information, i.e. millions of corner path failures. If so, address the cause of the excess data logging at the source.

    The database file and transaction log files do not shrink in physical size when they are archived;archiving only recovers space within the current physical size. As a result, insufficient archiving or leaving too much data in the database results in excessive growth over time.

    Russell Drown

Categories

In this Discussion