# Attach these extra labels to all timeseries collected by this Prometheus instance. Scrape_interval: 15s # By default, scrape targets every 15 seconds.Įvaluation_interval: 15s # Evaluate rules every 15 seconds. To enable the export of the metrics as well as some generic Tomcat export (for example around active threads) you need to comment in the following block in openmeetings/WEB-INF/web.xml There are canonical examples defined in the class definition Javadoc of the client packages.Io.Ġ.005,0.01,0.025,0.05,0.075,0.1,0.25,0.5,0.75,1,2.5,5,7.5,10Įnd Prometheus -> Enable Exposing the metrics and enable Tomcat generic metrics If you run Tomcat in embedded mode, please look at the AbstractTomcatMetricsTest for an example on how to configure the various exporters when running embedded. If you are defining your data source on application level (so inside your war), you need to set bindOnInit to ensure that your data source has been initialized before the metrics application starts.Enabling logFailed and logSlow may lead to a lot of additional metrics., so be careful !!!.slowQueryBuckets: the buckets separated by a pipe ("|") symbol to be used for the global query response times, defaults to 1|2.5|10|30.buckets: the buckets separated by a pipe ("|") symbol to be used for the global query response times, defaults to.threshold: the threshold in ms above which metrics will be collected if logSlow=true.logSlow: if set to 'true' collect metrics on metrics exceeding threshold.logFailed: if set to 'true' collect metrics on failed queries.A gauge with per query error counts (optional)Ĭonfiguration is usually done in Tomcat's server.xml or context.xmlĬonfiguration options of the interceptor are as shown above and have the following meaning:.A histogram with per query response times for slow queries (optional).A histogram with global query response times.The interceptor will collect the following metrics: Database response time metricsĭatabase response time metrics are only available when using the Tomcat JDBC Pool as this collector uses an interceptor mechanism that is only available for this type of pool. Nl.įor an explanation on histograms and buckets please see the prometheus documentation. There is no need to modify already deployed applications. If you want servlet response time metrics you can configure the TomcatServletMetricsFilter by adding it to the $CATALINA_BASE/conf/web.xml as shown below. RUN apt-get update & apt-get install -y curl & \Ĭurl -v -fail -location $.war -output /usr/local/tomcat/webapps/metrics.war The following Dockerfile provides an example how you include the exporter in a Tomcat image: After restart of tomcat you should be able to access metrics via the /metrics/ endpoint. Next, rename tomcat_exporter_servlet war file to metrics.war and add it to the webapps directory of Tomcat. Using the common.loader is important as we need to make sure that all metrics are registered using the same class loader. If you are running Tomcat in the conventional non-embedded setup we recommended to add the following jars (see pom.xml for the correct versions) to the $CATALINA_BASE/lib directory or another directory on the Tomcat common.loader path. The following Tomcat related metrics are provided: Prometheus Tomcat ExporterĪ set of collectors that can be used to monitor Apache Tomcat instances. I am considering sunsetting support for this exporter by the end of 2022. I suggest users to seriously consider to switch as you can expect more long term support from that implementation. NOTE: The official Prometheus java client now supports some overlapping functionality like the servlet filter.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |