Activity Observer

Owned by Jules DELEUZE (Unlicensed)
Last updated: Nov 13, 2019
4 min read

 

The activity consultation interface gives access to dashboards which include a number of statistical indicators on the instantaneous use of the engine since the last launch of Wedia.

Screen

General information

The upper part of the screen shows general information to know how long the server has been running, current memory usage according to the jvm, etc.

CPU

Two pieces of information are presented on the CPU:

  • system cpu : This is the percentage of CPU currently used on the entire server.
    So this brings together the activity of all the programs started on the server and not just the Wedia engine.

  • jvm cpu: this is the percentage of CPU currently used by Java jvm.
    So this brings together the activity of the application server, the Wedia engine, etc.

The marker board

The second part presents various markers (engine and application).

 

For each marker, it is possible to know:

  • Type: Specifies the name of the marker.
    In addition, the date below is the date on which the last action was taken.

  • In progress: how many calls are running at this time

  • Total: how many calls have been made since the server started. This figure also includes current applications.

  • Maximum in simultaneous: what is the maximum number of concurrent calls since starting the server.

  • Maximum duration: Indicates the time taken in milliseconds by the longest call since server startup for this marker.
    The date on which the appeal was made is also indicated.

  • Alerts: Some markers have pre-recorded alert thresholds.
    If the threshold is exceeded, we will know by this column the number of exceedances and the date of the last exceedance.

  • Average: a maximum sample of the last 200 calls is stored.
    From this sample, we present an average of the lead times, the minimum lead time, the maximum lead time, as well as the start and end of the sample.

  • History: The associated graph is the set of time delays of the sample for the marker.

Markers

  • Open DBMS connection: indicates all currently open connections. This represents the connections obtained and not closed.

  • Obtaining the connections DBMS: This counter measures only the connection to the DBMS and not its use.

It therefore represents the time it takes for the connection pool to connect to the motor.

This counter quite often shows that getting a connection can take a lot of time and this greatly affects the performance of the application.

The alert threshold has been set at 800 milliseconds, this time is largely sufficient for the connection pool to return an available connection.

If this time is long then it means that:

  • The connection pool is not sufficient: normally the maximum connection must be in adequacy with the number of http calls accepted by the application server.

  • The connection pool does not keep connections and wastes time opening connections with the DBMS: to avoid this, you only need to configure the connection minimum of the pool to the same value as the connection pool maximum.

Alerts and thresholds

When the duration of a marker exceeds the time indicated in the programmed "threshold", then the information is displayed on this screen.

In addition, a trace is recorded in the event log called "alert".
Thus the event log contains the alert history.

The trace also contains the whole stacktrace.

Sample Trace

10:52:02,110 WARN [alert] WEDIA ElasticSearch Watcher key:[pool_view_lib_access_sgbd_query_select] Sgbd query select duration: 3507 ms java.lang.RuntimeException: stacktrace at wsnoheto.engine.ObjectPoolManager.logAlerte(ObjectPoolManager.java:148) at wsnoheto.tools.Gate.stockDuration(Gate.java:421) at wsnoheto.tools.Gate._leave(Gate.java:396) at wsnoheto.tools.Gate.leave(Gate.java:371) at wsnoheto.tools.Gate.leave(Gate.java:355) at wsnoheto.database.baseDriver.getCardinality(baseDriver.java:2630) at wsnoheto.engine.CTObjectsLegacy.cardinality(CTObjectsLegacy.java:2770) at wsnoheto.index.elasticsearch.indexer.RunnableDeltaIndexer.<init>(RunnableDeltaIndexer.java:92) at wsnoheto.index.elasticsearch.ElasticSearchService$2.run(ElasticSearchService.java:380) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511) at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308) at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180) at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748)


Associated event log

The associated event log is named "alert".

It is accessible directly from the button at the top of the screen "Consult the logs of the thresholds".

Add an application marker

Learn how to add an application marker

Version

It is possible to capture the current state and make it a later viewable version.

  • This capture will contain all current page information.

  • It will be saved on the current server, under the name specified at the time of capture.

  • This information will be accessible later via the "Advanced Mode"button.

Advanced mode

Server versions

 

It is possible to

  • List the different captures made on the server.

  • Delete a version on the server.

  • Load a previous capture to view the activity consultation screen as it was at the time.
    In this case, the Activity View screen will display all information as it was at the time of the capture.

  • Recover a file representing a capture.
    This file can be sent by email to a developer so that they can display the page via another WEDIA engine, without having access to this functionality on this server.