UI Metrcis for pool's nodes

[Gaploid]

Describe the bug

1. You can't select 24 hours from the dropdown menu to show graphs of metrics. Its automatically switched to 7 days.
2. Network graph seems broken it shows strange numbers in petabytes.
   

To Reproduce
Steps to reproduce the behavior:

1. Go to pool
2. Click on dropdown menu "last" for metrics and select 24 hours
3. See that it will be switched to 7 days.

[timotheeguerin]

Fixing the dropdown issue. Could you try to run those 2 query directly in application insights and see what it returns you:
Just replace "insights-windows" with the name of your pool

customMetrics
| where timestamp >= ago(1h)
| where cloud_RoleName == "insights-windows"
| where name == "Network read"
| extend mb = value / 1000000000
| summarize avg(mb) by bin(timestamp, 5min), cloud_RoleInstance
| render timechart 

customMetrics
| where timestamp >= ago(1h)
| where cloud_RoleName == "insights-windows"
| where name == "Network write"
| extend mb = value / 1000000000
| summarize avg(mb) by bin(timestamp, 5min), cloud_RoleInstance
| render timechart 

This should show you a similar graph to what batch explorer has.

Also is this a windows or linux pool? This might be an issue with the data collected

[Gaploid]

ry to run those 2 query directly in application

Where exactly in-app insight can I execute these queries?

[timotheeguerin]

There is docs here https://github.com/Azure/batch-insights/wiki/Query-application-insights but basically go to the application insights in the azure portal and click on the analytics button

[timotheeguerin]

I'm going to close this as the first part is fixed and the second seems to be more of an issue with batch-insights.