| copyright | lastupdated | ||
|---|---|---|---|
|
2018-11-12 |
{:shortdesc: .shortdesc} {:new_window: target="_blank"} {:codeblock: .codeblock} {:pre: .pre} {:screen: .screen} {:tip: .tip} {:download: .download}
The load balancer monitors the following metrics:
- Throughput
- Active Connections
- Connection Rate
These metrics are visualized as graphs that can be viewed by selecting the Monitoring tab,
and are available programmatically as time series data points by using the getListenerTimeSeriesData API.
Each data point contains a timestamp in UNIX epoch time, and the metric value for that time interval ending at that timestamp. The user can specify the protocols and the time interval over which the metrics are to be reported.
Supported protocols include:
- HTTP
- HTTPS
- TCP
Specifying a protocol filters the metric by that protocol.
For example, if no protocol is specified, and the metric is Throughput, then the total throughput for all protocols is reported.
If the protocol is HTTP, then only the throughput for HTTP traffic is reported.
The user can also specify the time interval over which the metric is to be reported. The time intervals supported are:
- 1 hour
- 6 hours
- 12 hours
- 24 hours
- 1 week
- 2 weeks
The number of data points reported is roughly the same for each time interval.
For example, if the interval is 1 hour, then each data point represents 5 minutes of data.
If the interval is 2 weeks, then each data point represents 24 hours of data.
The table below shows how the data points are derived from the time interval:
| Time Interval | Datapoint Precision | Number of Datapoints |
|---|---|---|
| 1 hour | 5 minutes | 12 |
| 6 hours | 30 minutes | 12 |
| 12 hours | 1 hour | 12 |
| 24 hours | 2 hours | 12 |
| 1 week | 12 hours | 12 |
| 2 weeks | 24 hours | 12 |
In order to retrieve monitoring metrics you must link your SoftLayer account with your IBM© Cloud account. Refer to this topic 
for more information.