tibcosoftware / bw-tooling Goto Github PK

One of my team mate found out that,
we use 0.5.0 of java simpleclient prometheus lib
there already is 0.14.0
ours is from 2018
in 0.60 there was a memory leak
https://github.com/prometheus/client_java/releases?page=2

Thank you

We should provide built JAR file to be ready to use in our container base images.

Looks like source code is updated 2 years back and jar at location bw-tooling/prometheus-integration/sample/ is generated 3 years back.
It is helpful to keep the jar in sync with the code.

Thank you

I tried to monitor a BW AppNode with prometheus+grafana integration, and it works.
However, there are many cases when there are more than one AppNode on the same server, and only one of them can use
the metrics port (9095).
Is it possible to configure the metrics port (other than 9095) externally?

When service mesh (Linkerd) is used, Prometheus fails to get the metrics from Pod while giving EOF error.

No logs are printed in application or prometheus pod.

Following log is found in linkerd-proxy sidecar container -
hyper::proto::h1::role unexpected content-length found, canceling

Integrating Prometheus bundle with BWCE app causing a Memory leak. The BWCE pod/container not releasing memory and also not destroying a pod in case there is an OOM issue. You can check the below heap dump analysis screenshot :

JAR export finished with warnings. See details for additional information.
Exported with compile warnings: com.tibco.bw.prometheus.monitor/src/com/tibco/bw/prometheus/monitor/ProxyServlet.java
com.tibco.bw.prometheus.monitor/META-INF/MANIFEST.MF was replaced by the generated MANIFEST.MF and is no longer in the JAR.

Even though these are warnings I am getting some errors when I start docker container. I guess I missed this

This plugin provides a way to reset group variables at an application level rather than going to each and every process and specific activities to reset the variables.