![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
I am copying a subset of data from our production warehouse to a test bed warehouse. I am using Execute Package Tasks in which I place a Copy SQL Server Object Task. This task copies a single table's data replacing the existing data in the test bed table. I also include logging features so I can tell when the copy started, how long it took and whether it was successful. When I palce all the Execute Package Tasks in a single controling Execute Package Task, set the MaxConcurrentSteps to 4 and execute the controling task... some of the individual tasks fail telling me that "a file on the source server is being used by another user or process!". Additionally, to error IDs are embedded in the error message (80004005 and 80070020). I believe what is happening is that as each Copy SQL Server Object task is executed, a files on the source side are created which contain information about the tables, procedures, views indexes, etc that are to be copied. Although I can specify directory locations for logging and error files on the destination server (where the DTS Packages were created and are executed from) I can not specify the same on the Source Side. Nor can I specify that multiple Copy SQL Server Object Tasks are running and that DTS should allow for multiple packages needing access to "their own" versions of the work files. Does someone out there know how I can solve this? I have thought of using a Data Transformation Task with a Select * FROM Table (I actually use this for some of the tables and they do run concurrently) but I'm afraid this may take longer. Does it use BCP under the covers for this too. Some of my tables are 300+ Million records and I am worried I will not be able to accomplish the test bed rebuild in a single night (I must keep my test bed sychronized with other database I backup and restore to the test bed). Thanks |
![]() |
| Thread Tools | |
| Display Modes | |
| |