1 of 2

Monitoring

OpenWiFi 2.0 Telemetry and Analysis

TIP OpenWiFi software stack is envisioned to have a rich telemetry data that can be extracted, transformed and stored for analytics purposes. This section will outline various integration using the current capabilities of the OpenWiFi release. These integrations will provide examples for the community to enrich, adopt and productize.

The current release of OpenWiFi utilizes both a rich open API and Kafka for retrieving telemetry information from Access Points and SDK services. For the purpose of this section and Release 2.0 we will be showcasing Kafka integration with 3rd-party monitoring subsystems.

Kafka Data Source

The current release of 2.0 SDK architecture contains a Kafka broker for the purposes inter-services communication, state, healthcheck, device provisioning state producing and consuming Kafka topics. You can find the latest information related to Kafka topics here: https://github.com/Telecominfraproject/wlan-cloud-ucentralgw/blob/master/KAFKA.md#kafka-integration

The current Kafka topics used for this monitoring integration are:

state
healthcheck

All Kafka messages carry a JSON payload, example of a healthcheck message is as follow:

{
   "system":{
      "id":179033843641952,
      "host":"https://gw-ucentral-dev01.cicd.lab.wlan.tip.build:17002"
   },
   "payload":{
      "data":{
         "interfaces":{
            "up0v0":{
               "dhcp":false,
               "location":"/interfaces/0"
            }
         },
         "unit":{
            "memory":36
         }
      },
      "sanity":67,
      "serial":"112233445566",
      "uuid":1627357625
   }
}

A state Kafka message looks like:

{
   "system":{
      "id":179033843641952,
      "host":"https://gw-ucentral-dev01.cicd.lab.wlan.tip.build:17002"
   },
   "payload":{
      "serial":"112233445566",
      "state":{
         "interfaces":[
            {
               "clients":[
                  {
                     "ipv6_addresses":[
                        "fe80:0:0:0:206:aeff:fee0:69ad"
                     ],
                     "mac":"07:06:06:06:06:06",
                     "ports":[
                        "eth1"
                     ]
                  },
                  {
                     "ipv4_addresses":[
                        "192.168.4.1"
                     ],
                     "mac":"01:02:03:04:05:06",
                     "ports":[
                        "eth1"
                     ]
                  }
               ],
               "counters":{
                  "collisions":0,
                  "multicast":63,
                  "rx_bytes":14725,
                  "rx_dropped":0,
                  "rx_errors":0,
                  "rx_packets":209,
                  "tx_bytes":13571,
                  "tx_dropped":0,
                  "tx_errors":0,
                  "tx_packets":80
               },
               "dns_servers":[
                  "1.1.1.1",
                  "9.9.9.9"
               ],
               "ipv4":{
                  "addresses":[
                     "192.168.4.33/24"
                  ],
                  "leasetime":600
               },
               "location":"/interfaces/0",
               "name":"up0v0",
               "uptime":31349
            },
            {
               "counters":{
                  "collisions":0,
                  "multicast":0,
                  "rx_bytes":0,
                  "rx_dropped":0,
                  "rx_errors":0,
                  "rx_packets":0,
                  "tx_bytes":1058,
                  "tx_dropped":0,
                  "tx_errors":0,
                  "tx_packets":5
               },
               "ipv4":{
                  "addresses":[
                     "192.168.1.1/24"
                  ]
               },
               "location":"/interfaces/1",
               "name":"down1v0",
               "uptime":31355
            }
         ],
         "radios":[
            {
               "active_ms":24459917,
               "busy_ms":1173593,
               "channel":149,
               "channel_width":"80",
               "noise":4294967198,
               "phy":"soc/40000000.pci/pci0000:00/0000:00:00.0/0000:01:00.0",
               "receive_ms":4647,
               "transmit_ms":88272,
               "tx_power":30
            },
            {
               "active_ms":24456321,
               "busy_ms":11878205,
               "channel":11,
               "channel_width":"20",
               "noise":4294967204,
               "phy":"platform/soc/a000000.wifi",
               "receive_ms":1329,
               "transmit_ms":73228,
               "tx_power":30
            },
            {
               "active_ms":24458178,
               "busy_ms":1162312,
               "channel":36,
               "channel_width":"80",
               "noise":4294967192,
               "phy":"platform/soc/a800000.wifi",
               "receive_ms":12339,
               "transmit_ms":86904,
               "tx_power":23
            }
         ],
         "unit":{
            "load":[
               0.190921,
               0.263188,
               0.240726
            ],
            "localtime":1627418941,
            "memory":{
               "free":348540928,
               "total":520409088
            },
            "uptime":31386
         }
      },
      "uuid":1627357625
   }
}

ELK Integration

Kafka integration with ELK

The following pipeline is used to leverage Kafka messages being emitted from OpenWiFi 2.0 for ELK (Elastic Logstash Kibana) stack integration :

TIP OpenWiFi project has deployed an ELK stack for community members to access here.

The key for this integration is to use a plugin that enables Kafka to be used as an input for Logstash. This plugin can be found here. Once installed then Logstash can be configured to listen to the input source of the Kafka broker that is deployed as part of OpenWiFi SDK 2.0 release and its appropriate topics. Here is a sample Logstash configuration.

It is important to note that Logstash provides the ability to transform messages which then can be pushed to Elasticsearch for storage with effective indexing. Finally Kibana is used to create visualization such as this:

The following repository will be used to store necessary files for integration examples for monitoring.