What Is Monitored
How It Works
Deploy
KPIs
Overview
GermainUX reads the runtime state of each WebLogic Server over JMX and turns it into KPIs. You see the execute thread pool with its hogging and stuck threads, every JDBC data source and how many requests wait for a connection, the depth of each JMS destination, and transaction commits and rollbacks. Thirty-one KPIs, default SLAs on server state and health, and a Weblogic Overview dashboard come ready to use. Your team finds the server, data source or queue behind a slow application without opening the WebLogic console on each server.Stuck and Hogging Threads
JDBC Pool Pressure
JMS Backlogs
Server State and Health
RUNNING and reports HEALTH_OK, with a default SLA on both.Questions It Answers
- Which WebLogic server has stuck threads right now, and since when?
- Is a JDBC data source out of connections, and are requests failing to get one?
- Is a JMS queue backing up because its consumers stopped?
- Are more transactions rolling back than usual?
- Which servlets are slowest, and in which application?
- Is every server running and healthy?
- Did a data source stop running, for example during a database failover?
What Is Monitored
Analytics, Alerts and Automation
AI Analysis
Ask AI Explore questions such as “Which WebLogic data sources had requests waiting for a connection this week?” Outlier detection flags a thread count, queue depth or rollback count outside its normal range.Dashboards and Reports
In the left menu, go to Dashboards > All, and then open the Weblogic Overview dashboard. It has one chart or table for each of the 31 WebLogic KPIs: trend charts for threads, transactions and JDBC waits, and tables for data sources, JMS destinations, server state, servlets and work managers. Click a chart to open its KPI and break it down by server or by name. You can build your own dashboards and reports on the same KPIs.
Weblogic Overview Dashboard - GermainUX
Analysis
Click a value on the dashboard to open it in the Analysis screen. It compares the selected period with earlier ones, and the breakdown shows which server, data source or destination changed most.
Analysis of WebLogic Stuck Threads - GermainUX
Investigate an Exhausted JDBC Pool
When requests slow down or fail to reach the database, open the WebLogic JDBC Waiting For Connection Current Count KPI in the Analysis screen. The top row compares the average number of waiting requests with the previous days and shows the hourly trend. The relationships chart shows when the requests started to wait, and the segmentation breakdown lists what changed most. Then open WebLogic JDBC Waiting For Connection Failure Total and WebLogic JTA Transaction Rolled Back Total Count to see whether requests gave up and transactions rolled back. In the example below, a batch job locks the orders tables, and theOrdersDS data source runs out of connections on orders_ms1 and orders_ms2. On average, 3.889 requests wait for a connection over six hours, against 0 on the previous days. The relationships chart peaks at about 20 waiting requests, and the breakdown marks the Order Management application as a spike. The requests stop waiting about 50 minutes later, when the batch job releases its locks. Add an SLA on this KPI to get an alert as soon as requests start to wait.

Analysis of WebLogic JDBC Waiting For Connection Current Count - GermainUX
Investigate a JMS Backlog
When messages stop moving through a queue, open the WebLogic JMS Messages Current Count KPI in the Analysis screen. The top row compares the average queue depth with the previous days, and the segmentation breakdown lists the servers whose destinations grew most. Then check WebLogic JMS Consumers Current Count for the same destination: a destination without consumers keeps growing. In the example below, the fulfilment system is down and theFulfilmentMDB message-driven bean is suspended. OrderQueue has no consumers on either server for 100 minutes, and the messages pile up. The average depth is 118.111 messages over six hours, against 4.943 on the previous days, and the relationships chart peaks at about 400 messages. The breakdown marks both servers as spikes, wls-01 with 121.861 messages and wls-02 with 114.361. The queue drains once the consumers come back.

Analysis of WebLogic JMS Messages Current Count - GermainUX
Investigate a Failed Managed Server
When a managed server stops serving requests, open the WebLogic Server State KPI in the Analysis screen. Each reading is named after the state the server reports, so the segmentation breakdown lists the states that changed most. The top row compares the average with the previous days and shows the hourly trend. In the example below,orders_ms1 runs out of memory during a nightly export, fails, and Node Manager restarts it. The average state is 97.101% over six hours, against 100.000% on the previous days. The monitor records 69 readings instead of 72, because orders_ms1 doesn’t answer while it’s down. The breakdown marks ADMIN, a state orders_ms1 reports on its way back, as new and as an outlier. The state returns to 100% once the server is RUNNING again.

Analysis of WebLogic Server State - GermainUX
Investigate a Suspended Data Source
When applications lose their database on every server at once, open the WebLogic JDBC State KPI in the Analysis screen. The KPI is100 while a data source is running and 0 otherwise, so the average drops as soon as one stops. The top row compares the average with the previous days and shows the hourly trend, and the relationships chart shows when the data source stopped and came back.
In the example below, the database fails over, and WebLogic suspends OrdersDS on orders_ms1 and orders_ms2 for about half an hour. The average state is 90.000% over five hours, against 100.000% on the previous days, and the relationships chart dips to 50%. The breakdown finds no leading factor, because the data source stops on both servers at the same time. The state returns to 100% once OrdersDS runs again. Add an SLA on this KPI to get an alert when a data source stops running.

Analysis of WebLogic JDBC State - GermainUX
SLAs and Alerts
WebLogic Server State and WebLogic Server Health State come with a default SLA that is breached when a server isn’tRUNNING or doesn’t report HEALTH_OK. Add SLAs on other KPIs, such as stuck threads or requests waiting for a connection, and get an alert by email, SMS, Slack or Teams when one is breached.
Automation
When an SLA is breached, an automated action can run a diagnostic script, such as a thread dump, call an HTTP endpoint or open an incident.Correlation with Other Technologies
Correlate WebLogic data with the Oracle Database behind your data sources and with the browser sessions of your users, to follow a slow page to the pool or queue that caused it.How It Works
Before You Start
- A GermainUX Engine can reach the listen port of every WebLogic server you monitor.
- You know the JMX service URL of the runtime MBean server of each WebLogic server.
- You have a WebLogic user for the monitor, for example one in the
Monitorsgroup, which has read-only access to the MBeans. - Firewalls allow the Engine to reach the listen ports.
Deploy
Open the Java Monitor Deployment wizard
Choose the WebLogic monitor

Java Monitor Deployment Wizard, Monitoring Step - GermainUX
Connect to the server
Deploy it to the Engine
WebLogic orders_ms1, and select the Monitoring Node and Engine. Set the schedule: the default, 0 0 * * * *, runs once an hour, and 0 */5 * * * * reads the server every 5 minutes. Finish the wizard.Repeat for the other servers
Validate
Wait for the first run
0 */5 * * * *.Open the WebLogic dashboard
Check server state
RUNNING, and WebLogic Server Health State as HEALTH_OK.Troubleshooting
No data appears for a server
No data appears for a server
The connection fails with an authentication error
The connection fails with an authentication error
Monitors or Administrators group.JDBC or JMS tables are empty
JDBC or JMS tables are empty
Security and Privacy
Oracle WebLogic monitoring data includes:- System data: the host names of the WebLogic servers and the names of the servers
- Application data: the names of your applications, web modules, servlets, data sources, JMS destinations and work managers
KPIs
100 or higher, per server host. The other KPIs have no default SLA. Add the thresholds you need. See KPIs and SLAs.
Related Documentation
Java Monitoring
Oracle DB Observability
Automation
Get Help
The GermainUX team can help you monitor other WebLogic MBeans, such as clusters, EJB pools or the Administration Server’s domain runtime.Contact GermainUX Support
Can't Find What You Need?
The content of this document is proprietary to GERMAIN SOFTWARE.

