Welcome, Phillip. I noticed a curious thing the other day. One of my servers, which is crowded with jobs, was running slower than usual. I took a pen and paper and created a list of start and stop times using it to reorganize the schedule and minimize overlaps. I'd hoped for some mild performance increase but was surprised to see 30-35% improvement. Why was this so dramatic? What can I do to eliminate the manual effort of tracking the job schedule on paper?

    Requires Free Membership to View

    Login

This is hard to really say for sure without seeing the top wait events during the timeframe in question. However, this was probably due to resource contention especially if any of these jobs update any data. If process 1 has a resource locked that process 2 needs, then process 2 has to wait on process 1 to finish and free the lock. The more processes updating the same table, the more likely this will occur.

This was first published in May 2007

Join the conversationComment

Share
Comments

    Results

    Contribute to the conversation

    All fields are required. Comments will appear at the bottom of the article.
    Back to top