January 12, 2006 at 2:40 pm
I developed a package on an XP platform that accesses a remove SQL Server (for both source and destination connections) that is hosted on an old, single core, single CPU with 1/2 GB memory and the execution time is 1:47.532. I move the package to the target machine which has dual CPUs and 4 GB of RAM (and NO other users, applications, or additional processes) and the execution duration is 3:56.110. When testing I am running it out of Visual Studio 2005. When moved to the target machine I am running the bin and editing the four connection managers before exection. Otherwise the tables and data are identical, and the resultant rows in each of the tables is the same. I have disconnected the test machine from the network to confirm that I am not somehow inadvertently connecting back to the development machine. The target machine has hyper-threading turned off and AWE turned on. Any ideas as to why it is so slow?
January 16, 2006 at 8:00 am
This was removed by the editor as SPAM
January 16, 2006 at 2:57 pm
Now, as this package has gotten progressively larger the statistics are a bit different. The same two hosts, a low-level AMD with 1/2 GB RAM for development, and a pair of high-level P4s with 4 GB RAM for production. Using dtexecUI to run against both machines results in times of 6:16.127 and 6:37.684 respectively. Running the same package from Visual Studio on an XT client and the stat posted in the progress tab (window, sheet) is 1:52.328.
Why does it take four times longer to run in the exec utility than in the GUI? Why are the times so similar between a crippled machine and a powerful host?
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply