Uploaded image for project: 'Confluence Data Center'
  1. Confluence Data Center
  2. CONFSERVER-12837

Provide recommendations for caches in the UI

    XMLWordPrintable

Details

    • We collect Confluence feedback from various sources, and we evaluate what we've collected when planning our product roadmap. To understand how this piece of feedback will be reviewed, see our Implementation of New Features Policy.

    Description

      Cache tuning should be more automated; when a cache size needs to increase, the cache admin page should provide a recommended value (and an easy way to enter it).

      Atlassian Status as of 18 Oct 2010

      Hi All,
      Thanks for your votes for this issue. The Confluence team had spent some time attempting to provide a recommended cache setting in the admin user interface. After careful consideration, we have come to the conclusion that this is something we cannot implement. Despite our efforts, we came to understand there is no cheep, automated, accurate method to provide cache size recommendations.

      There are several factors that input into providing an effective cache recommendation strategy. These are:

      • Memory size allocation
        Simply increasing the cache size could lead to more memory usage and an increase in full garbage collections performed by the Java Virtual Machine (JVM). This could consequently lead to even worse performance.
      • Selective cache tuning
        Some memory caches are more important than others. Changing those that are used more extensively would provide more benefits. This should be done through a manual tuning process.
      • Other cache considerations
        • Size of an object stored inside a cache should be considered when recommending the size.
        • Cache entries expiry time.
      • The tuning process
        When changing cache it is important to check how the change affected hit/miss/usage - It's important to keep an eye on the performance of the application after the cache sizes have been changed (see related documentation) - there is no cheap way to automate this.

      We recommend that all customers spend some time going through our Cache Performance Tuning documentation and conduct a manual tuning process - there is no hard and fast rule to solve this problem.

      Sherif Mansour
      Atlassian Product Management

      Attachments

        Issue Links

          Activity

            People

              merickson Matthew Erickson
              jlargman Jeremy Largman
              Votes:
              42 Vote for this issue
              Watchers:
              18 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: