Have you ever wondered how many CPU seconds it takes to probe an instance via TCP or HTTP 100, 1,000, or 10,000 times?
In this blogpost we'll be taking a look at how many CPU seconds in total are used across four cores when probing an external service using the blackbox exporter.
In case you are unfamiliar with the concept, blackbox monitoring is the monitoring of a service whereby no participation of the monitored system is required. The monitoring only observes external functionality; essentially what the outside world sees.
The blackbox exporter provided by the Prometheus organisation allows for blackbox probing of endpoints over HTTP, HTTPS, DNS, TCP, and ICMP.
For this test I used the Intel Core i7-4980HQ CPU.
The following results were recorded when calculating the average time it took for the blackbox exporter to send X amount of TCP/HTTP probes to a locally running Prometheus instance:
TCP Probes:
No. of probes
100
1,000
10,000
Time (s)
0.07s
0.81s
7.85s
HTTP Probes:
No. of probes
100
1,000
10,000
Time (s)
0.09s
1.01s
9.6s
As we can see, it is possible to do 1,000 probes per second with relatively little cost in regards to CPU usage.
The results of these probes can be scraped by a Prometheus instance, queried, graphed, or turned into alerts.
Have questions about how to blackbox probe your network? Contact us.
No comments.