Active Network Testing

NETPROBE Handler

DOCUMENT STATUS: DRAFT

The idea of a netprobe handler follows the learnings of the September 2022 Hackathon, where Cloudprober was successfully implemented as a new back-end to the Orb agent. Though functional, the cloudprober approach had a few shortcomings, most notably in the metrics it provided. An alternate approach using pktvisor is the intent behing the netprobe handler.

Supported Test Types

The netprobe handler should support multiple test types and should be extensible to add additional test types in the future. For an initial release, supporting the PING (ICMP Echo) test type should be sufficient.

The following test types should be considered for future development:

• PING with Jitter measurements
• TCP Connect
• HTTP/HTTPS
• UDP Echo

Supported Metrics

The following set of metrics should be supported across all test types:

Metric	Type	Unit of Measure	Metric Group	Description
attempts	counter	integer	Counters	Number of times a test was run in the reporting interval
successes	counter	integer	Counters	Number of successful tests in the reporting interval
total_time	counter	time (in microseconds?)	Counters	Cumulative elapsed time for all successful tests in the reporting interval
response_max	gauge	time (in microseconds?)	Counters	Maximum response time measured in the reporting interval
response_min	gauge	time (in microseconds?)	Counters	Minimum response time measured in the reporting interval
response_quantiles	quantiles	time (in microseconds?)	Quantiles	Quantiles of test response times (P50/90/95/99)

Proposed Policy Format

Here are a few concepts to keep in mind in reviewing this:

• the netprobe tap provides the basic facilities to run network tests, but should only include host specific configuration settings and default configuration overrides
• the netprobe input is where the network tests are defined and configured, as well as any default configuration overrides
• the netprobe handler is where the metrics that are to be measured and collected are configured

Here is a sample configuration for the netprobe tap:

version: "1.0"

visor:
  taps:
    default_netprobe:
      input_type: netprobe
      config:
        maximum_concurrent_tests: 10
        ip_source_binding: 127.0.0.1
      tags:
        virtual: true
        vhost: 1 

Here is a sample policy for the netprobe input and handler:

version: "1.0"

visor:
  policies:
    basic_ping_policy:
      kind: collection
      description: "basic PING netprobe policy"
      input:
        tap: default_netprobe
        input_type: netprobe
        config:
          test_type: ping
          interval_msec: 2000
          timeout_msec: 1000
          packets_per_test: 10
          packets_interval_msec: 25
          packet_payload_size: 56
          disable_scout_packet: false
          disable_integrity_check: false
          targets:
            test_1_name:
              target: foo.bar
            test_2_name:
              target: 10.0.0.1
              tos: EF
      handlers:
        config:
          num_periods: 2 #default is 5
        modules:
          default_ping:
            type: netprobe
            config:
              latency_units: msec
            metric_groups:
              enable:
                - quantiles
                - dns_resolution
              disable:
                - jitter