The PR is #4008

Last time I tried to get this working, the issue was with getting python.d
to properly load the Yoctopuce modules. As far as I could tell at the time, the dynamic import thing works, it's just an issue with getting Python to properly import modules that aren't in system.path, but I didn't look very deep beyond that. Part of the problem there is that the Yoctopuce Python

Kibana 8.0

The roles when added to the user, look smudged- Plz introduce a space or a comma to make it more readable.
Here there are 3 roles- test_logstash_reader, reporting_user and global_dashboard_read. But looks crammed:

<img width="946" alt="Screen Shot 2020-09-03 at 12 39 40 PM" src="https://user-images.githubusercontent.com/11466284/92176328-16963500-edf3-11ea-834d-01ca637d2534.png

Please answer these questions before submitting your issue.

• Why do you submit this issue?
• [] Question or discussion
• Bug
• Requirement
• Feature or performance improvement

Question

org.springframework.kafka.KafkaListenerEndpointContainer#0-4-C-1 InstMethodsInter : class[class org.postgresql.jdbc.PgPreparedStatement] before method[executeWithFlags] intercept fai

This is a FEATURE REQUEST.

Please add support for kube_state_metrics_build_info as part of the metrics exposed on telemetry port 8081.
kube_state_metrics_build_info shall include the version of the kube-state-metrics.
Most Prometheus exporters are currently supporting such metric.

Is your feature request related to a problem? Please describe.
I use a windows computer develop kiali/ui but the run script don't run

Describe the solution you'd like
on windows platform can develop ui project.

Describe alternatives you've considered
can add cross-env to fix cross platform problem

Additional context
This is a great project

Cloudprober supports building additional metrics (other than the default ones) from external probe output. We could possible do the same for HTTP probe.

Why: When opening the Event Injection modal for a particular eventBus, the tool populates each field (source, detail type, and detail) with the last values used for that eventBus. Allowing all events to be saved, similar to the command history in the terminal, will allow users to quickly test multiple events without having to type the same parameters repeatedly.

Expected Behaviour:

Currently, JavaScript SDK logs status code 0 when the connection between the client and server cannot be resolved. This makes diagnosing network errors very difficult, as there are many different problems that can cause 0 status code in failing dependency calls.

The proposal is to log status code information that indicates whether a network error occurred and what type.

Describe the bug
Deploying a new job, then pulling logs via the CLI results in a panic (SIGSEGV).

To Reproduce

1. odin deploy -f test.job.yaml
2. odin log -i {job_id}

vagrant@vagrant:~/odin$ odin log -i dea194794589
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x40 pc=0x79c764]

goroutine

Has anyone done this safely?