Uploaded image for project: 'Nuxeo Platform'
  1. Nuxeo Platform
  2. NXP-22994

Reduce number of metrics published to Graphite

    XMLWordPrintable

    Details

    • Type: Improvement
    • Status: Resolved
    • Priority: Major
    • Resolution: Fixed
    • Affects Version/s: None
    • Fix Version/s: 9.3
    • Component/s: Core
    • Release Notes Description:
      Hide

      A mechanism to authorize some metrics and to filter some from being sent to graphite has been added.
      <extension target="org.nuxeo.runtime.metrics.MetricsService" point="configuration">
      <graphiteReporter enabled="true" periodInSecond="20" host="graphite.server.name" port="2003"
      prefix="servers.${hostname}.nuxeo.">
      <allowedMetrics>
      <metric>nuxeo.cache.user-entry-cache</metric>
      <metric>nuxeo.cache.group-entry-cache</metric>
      </allowedMetrics>
      <deniedMetrics>
      <metric>nuxeo.cache</metric>
      <metric>nuxeo.directories</metric>
      </deniedMetrics>
      </graphiteReporter>
      </extension>

      Show
      A mechanism to authorize some metrics and to filter some from being sent to graphite has been added. <extension target="org.nuxeo.runtime.metrics.MetricsService" point="configuration"> <graphiteReporter enabled="true" periodInSecond="20" host="graphite.server.name" port="2003" prefix="servers.${hostname}.nuxeo."> <allowedMetrics> <metric>nuxeo.cache.user-entry-cache</metric> <metric>nuxeo.cache.group-entry-cache</metric> </allowedMetrics> <deniedMetrics> <metric>nuxeo.cache</metric> <metric>nuxeo.directories</metric> </deniedMetrics> </graphiteReporter> </extension>
    • Epic Link:
    • Tags:
    • Sprint:
      nxcore 9.3.5, nxcore 9.3.6, nxcore 9.3.7
    • Story Points:
      3

      Description

      Metrics used to be accessible in JMX and published in Graphite.

      Since the directories and the new cache metrics has been added there is way too much metrics (~800 on a default instance) to report all of them through Graphite, instead of filtering at the Graphite level, it is better to not collect them and not send them in the metric GraphiteReporter.

      Having so much metrics impacts application perfs (time to collect them every metrics.graphite.period) and overload the graphite server disk IO especially in the case of a Nuxeo cluster.

      A filter should be added to limit the number of metrics published.

        Attachments

          Issue Links

          is related to

          Improvement - An improvement or enhancement to an existing feature or task. NXP-20628 Caches should have metrics by default

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Resolved

            Activity

              People

              • Votes:
                0 Vote for this issue
                Watchers:
                2 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved:

                  Time Tracking

                  Estimated:
                  Original Estimate - 0 minutes
                  0m
                  Remaining:
                  Remaining Estimate - 0 minutes
                  0m
                  Logged:
                  Time Spent - 2 days
                  2d