Forum Discussion
Azure VM monitor
For 1,2 and 3 Metric alerts are on the fastest pipeline, so you will see them quickly - subject to being able to set Guest metric alerts in your case. However in all three cases being able to view and query the data in log Analytics can also provide benefits. E,g, Alerts just look at the past 24hrs, so you will miss patterns that occur beyond that, a Log query can go back and look at whatever data you have retained.
Go to Log Analytics and Run Query
Sample output:
| CounterName | TimeGenerated | avg_CounterValue |
|---|---|---|
| % Processor Time | 2019-06-04T05:00:00Z | 60.91787148115488 |
| % Used Memory | 2019-06-04T05:00:00Z | 31.629809780248415 |
| % Free Space | 2019-06-04T05:00:00Z | 82.89545989470517 |
| % Free Space | 2019-06-04T02:00:00Z | 82.91151748380445 |
| % Processor Time | 2019-06-04T02:00:00Z | 59.627356628337175 |
| % Used Memory | 2019-06-04T02:00:00Z | 31.471773493138787 |
| % Processor Time | 2019-06-04T03:00:00Z | 59.52993177313613 |
| % Free Space | 2019-06-04T03:00:00Z | 82.8896254469986 |
| % Used Memory | 2019-06-04T03:00:00Z | 31.54788573413215 |
| % Processor Time | 2019-06-04T04:00:00Z | 59.35017016026048 |
Neither do methods do #4 well, there are some data points in the Log, such as Heartbeat but its not a reliable up/down indicator on its own. e.g. the Heartbeat can fail but the server is still up. You can check certain EventIDs, but if you get a crash and the server never comes back up, you might not see what the last EventID was.
Queries that can show Heartbeat are:
Go to Log Analytics and Run Query
Go to Log Analytics and Run Query
More examples are shown in the LOGS portal, when you open a new Query tab.