Performance Metrics

Since v6.1.0/31

The Performance Metrics represent system-level performance measurements for machines running the HVR Hub (hub) and HVR Agent (agent). These metrics are averaged on a minute-by-minute basis and include CPU usage, disk activity, and network IO. They are supported for both Capture and Integrate jobs.

Here is the list of Performance Metrics.

Performance Metric	Description
CPU Usage	Average CPU usage of a hub or agent machine per minute, represented in percentage.
IO Wait	The time spent by CPU of a hub or agent machine waiting for IO, represented in percentage.
Network Dropped Incoming/Outgoing Packages	The number of dropped incoming/outgoing network packages of a hub or agent machine.
Network Incoming/Outgoing Errors	The number of incoming/outgoing network errors of a hub or agent machine.
Config/Temp/HVR_HOME Disk Time Spent Reading/Writing	The time the disk spent reading/writing to the HVR_HOME, HVR_CONFIG, and TEMP directories on a hub or agent machine, measured in milliseconds.
Config/Temp/HVR_HOME Disk Time Spent doing IO	The time the disk spent on IO operations for the HVR_HOME, HVR_CONFIG, and TEMP directories on a hub or agent machine, measured in milliseconds.
Config/Temp/HVR_HOME Disk Weighted Time Spent doing IO	The weighted time the disk spent on IO operations for the HVR_HOME, HVR_CONFIG, and TEMP directories on a hub or agent machine in milliseconds.
Config/Temp/HVR_HOME Disk IO Operations In Progress	The number of in-progress disk IO operations for the HVR_HOME, HVR_CONFIG, and TEMP directories on a hub or agent machine.

Enabling Performance Metrics

The Performance Metrics are disabled by default. To enable them, set the environment variable HVR_PERFORMANCE_METRIC_ENABLE to 1 for the hub and agent.

To set HVR_PERFORMANCE_METRIC_ENABLE on the hub, add the Job System Environment Variable to your hub. You can also set the environment variable using the command hvrjobconfig with option -E.
To set HVR_PERFORMANCE_METRIC_ENABLE on the source and target, add the Environment action to both locations.

Disabling Performance Metrics

To disable Performance Metrics, set the environment variable HVR_PERFORMANCE_METRIC_ENABLE to 0 on the hub, source, and target.

Performance Metrics Log Examples

When Performance Metrics are enabled, the measures are represented in the logs for each category as follows. Logs from the hub are prefixed with <Hub>, and logs from the agent are prefixed with <Agent>.

CPU Usage In the following example, the average CPU Usage of machine running the hub 5%.
```
<Hub> Average CPU usage of machine 'demo' for the last minute is 5%.
```
I/O Wait
In the following example, the time spent by the CPU of the agent machine waiting for IO is 4.40%.
```
<Agent> Time spent by the CPU waiting for a IO on machine 'demo' is 4.40%.
```
Network Incoming/Outgoing Errors
In the following example, the machine running the agent reported a number of dropped incoming packets, dropped outgoing packets, incoming errors, and outgoing errors.
```
<Agent> Network interface em1 in machine 'demo' dropped 1 incoming packet and 5 outgoing packets. It also had 3 incoming errors and 6 outgoing errors.
```

Config/Temp/HVR_HOME Disk IO Operations In Progress

In this example, the machine running the hub reported the time spent in the HVR_CONFIG directory.

<Hub> Disk config partition dm-2 (used by directory /home/user1/dev/hvr_observability/hvr_config) in machine 'demo' spent 118742472 ms reading, 2834220646 ms writing and 52780841 ms doing IO (weighted: 2953464018 ms). With 0 IO operations in progress.