17. Statistics

17.1. Statistics Overview

Both Kea DHCP servers support statistics gathering. A working DHCP server encounters various events that can cause certain statistics to be collected. For example, a DHCPv4 server may receive a packet (the pkt4-received statistic increases by one) that after parsing is identified as a DHCPDISCOVER (pkt4-discover-received). The server processes it and decides to send a DHCPOFFER representing its answer (the pkt4-offer-sent and pkt4-sent statistics increase by one). Such events happen frequently, so it is not uncommon for the statistics to have values in the high thousands. They can serve as an easy and powerful tool for observing a server’s and a network’s health. For example, if the pkt4-received statistic stops growing, it means that the clients’ packets are not reaching the server.

There are four types of statistics:

  • integer - this is the most common type. It is implemented as a 64-bit integer (int64_t in C++), so it can hold any value between -2^63 to 2^63-1.
  • floating point - this type is intended to store floating-point precision. It is implemented as a C++ double type.
  • duration - this type is intended for recording time periods. It uses the `boost::posix_time::time_duration type, which stores hours, minutes, seconds, and microseconds.
  • string - this type is intended for recording statistics in textual form. It uses the C++ std::string type.

During normal operation, the DHCPv4 and DHCPv6 servers gather statistics. For a list of DHCPv4 and DHCPv6 statistics, see Statistics in the DHCPv4 Server and Statistics in the DHCPv6 Server, respectively.

To extract data from the statistics module, the control channel can be used. See Management API for details. It is possible to retrieve a single statistic or all statistics, reset statistics (i.e. set to a neutral value, typically zero), or even completely remove a single statistic or all statistics. See the section Commands for Manipulating Statistics for a list of statistics-oriented commands.

Statistics can be used by external tools to monitor Kea. One example of such a tool is Stork. See Monitoring Kea with Stork for details how to use it to retrieve statistics periodically (and use other data sources) to get better insight into Kea health and operational status.

17.2. Statistics Lifecycle

In Kea 1.6.0 version and earlier, when the Kea server is started some of the statistics are initially not initialized. For example, the pkt4-received statistic is not available until the first DHCP packet is received. In the later Kea versions, this behavior has been changed and all of the statistics supported by the servers are initialized upon the servers’ startup and should be returned in response to the commands such as statistic-get-all. The runtime statistics concerning DHCP packets processed is initially set to 0 and is reset upon the server restart.

Per-subnet statistics are recalculated when reconfiguration takes place.

In general, once a statistic is initialized it is held in the manager until explicitly removed, by statistic-remove or statistic-remove-all being called, or when the server is shut down.

Removing a statistic that is updated frequently makes little sense, as it will be re-added when the server code next records that statistic. The statistic-remove and statistic-remove-all commands are intended to remove statistics that are not expected to be observed in the near future. For example, a misconfigured device in a network may cause clients to report duplicate addresses, so the server will report increasing values of pkt4-decline-received. Once the problem is found and the device is removed, the system administrator may want to remove the pkt4-decline-received statistic, so it will not be reported anymore. If a duplicate address is ever detected again, the server will add this statistic back.

17.3. Commands for Manipulating Statistics

There are several commands defined that can be used for accessing (-get), resetting to zero or a neutral value (-reset), or removing a statistic completely (-remove). We can change the statistics time based limit (-sample-age-set) and size based limit (-sample-count-set) which control how long or how many samples of the given statistic are retained.

The difference between reset and remove is somewhat subtle. The reset command sets the value of the statistic to zero or a neutral value, so after this operation, the statistic will have a value of 0 (integer), 0.0 (float), 0h0m0s0us (duration), or “” (string). When requested, a statistic with the values mentioned will be returned. Remove removes a statistic completely, so the statistic will no longer be reported. Please note that the server code may add it back if there is a reason to record it.

Note

The following sections describe commands that can be sent to the server; the examples are not fragments of a configuration file. For more information on sending commands to Kea, see Management API.

17.3.1. The statistic-get Command

The statistic-get command retrieves a single statistic. It takes a single-string parameter called name, which specifies the statistic name. An example command may look like this:

{
    "command": "statistic-get",
    "arguments": {
        "name": "pkt4-received"
    }
}

The server returns details of the requested statistic, with a result of 0 indicating success and the specified statistic as the value of the “arguments” parameter. If the requested statistic is not found, the response will contain an empty map, i.e. only { } as an argument, but the status code will still indicate success (0). An example response:

{
    "command": "statistic-get",
    "arguments": {
        "pkt4-received": [ [ 125, "2019-07-30 10:11:19.498739" ], [ 100, "2019-07-30 10:11:19.498662" ] ]
    },
    "result": 0
}

17.3.2. The statistic-reset Command

The statistic-reset command sets the specified statistic to its neutral value: 0 for integer, 0.0 for float, 0h0m0s0us for time duration, and “” for string type. It takes a single-string parameter called name, which specifies the statistic name. An example command may look like this:

{
    "command": "statistic-reset",
    "arguments": {
        "name": "pkt4-received"
    }
}

If the specific statistic is found and the reset is successful, the server responds with a status of 0, indicating success, and an empty parameters field. If an error is encountered (e.g. the requested statistic was not found), the server returns a status code of 1 (error) and the text field contains the error description.

17.3.3. The statistic-remove Command

The statistic-remove command attempts to delete a single statistic. It takes a single-string parameter called name, which specifies the statistic name. An example command may look like this:

{
    "command": "statistic-remove",
    "arguments": {
        "name": "pkt4-received"
    }
}

If the specific statistic is found and its removal is successful, the server responds with a status of 0, indicating success, and an empty parameters field. If an error is encountered (e.g. the requested statistic was not found), the server returns a status code of 1 (error) and the text field contains the error description.

17.3.4. The statistic-get-all Command

The statistic-get-all command retrieves all statistics recorded. An example command may look like this:

{
    "command": "statistic-get-all",
    "arguments": { }
}

The server responds with details of all recorded statistics, with a result set to 0 to indicate that it iterated over all statistics (even when the total number of statistics is zero). An example response returning all collected statistics:

{
    "command": "statistic-get-all",
    "arguments": {
        "cumulative-assigned-addresses": [ [ 0, "2019-07-30 10:04:28.386740" ] ],
        "declined-addresses": [ [ 0, "2019-07-30 10:04:28.386733" ] ],
        "reclaimed-declined-addresses": [ [ 0, "2019-07-30 10:04:28.386735" ] ],
        "reclaimed-leases": [ [ 0, "2019-07-30 10:04:28.386736" ] ],
        "subnet[1].assigned-addresses": [ [ 0, "2019-07-30 10:04:28.386740" ] ],
        "subnet[1].cumulative-assigned-addresses": [ [ 0, "2019-07-30 10:04:28.386740" ] ],
        "subnet[1].declined-addresses": [ [ 0, "2019-07-30 10:04:28.386743" ] ],
        "subnet[1].reclaimed-declined-addresses": [ [ 0, "2019-07-30 10:04:28.386745" ] ],
        "subnet[1].reclaimed-leases": [ [ 0, "2019-07-30 10:04:28.386747" ] ],
        "subnet[1].total-addresses": [ [ 200, "2019-07-30 10:04:28.386719" ] ]
    },
    "result": 0
}

17.3.5. The statistic-reset-all Command

The statistic-reset command sets all statistics to their neutral values: 0 for integer, 0.0 for float, 0h0m0s0us for time duration, and “” for string type. An example command may look like this:

{
    "command": "statistic-reset-all",
    "arguments": { }
}

If the operation is successful, the server responds with a status of 0, indicating success, and an empty parameters field. If an error is encountered, the server returns a status code of 1 (error) and the text field contains the error description.

17.3.6. The statistic-remove-all Command

The statistic-remove-all command attempts to delete all statistics. An example command may look like this:

{
    "command": "statistic-remove-all",
    "arguments": { }
}

If the removal of all statistics is successful, the server responds with a status of 0, indicating success, and an empty parameters field. If an error is encountered, the server returns a status code of 1 (error) and the text field contains the error description.

17.3.7. The statistic-sample-age-set Command

The statistic-sample-age-set command sets time based limit for collecting samples for given statistic. It takes two parameters a string called name, which specifies the statistic name and integer value called duration, which specifies the time limit for given statistic in seconds. An example command may look like this:

{
    "command": "statistic-sample-age-set",
    "arguments": {
        "name": "pkt4-received",
        "duration": 1245
    }

}

The server will respond with message about successfully set limit for the given statistic, with a result set to 0 indicating success and an empty parameters field. If an error is encountered (e.g. the requested statistic was not found), the server returns a status code of 1 (error) and the text field contains the error description.

17.3.8. The statistic-sample-age-set-all Command

The statistic-sample-age-set-all command sets time based limits for collecting samples for all statistics. It takes single-integer parameter called duration, which specifies the time limit for given statistic in seconds. An example command may look like this:

{
    "command": "statistic-sample-age-set-all",
    "arguments": {
        "duration": 1245
    }

}

The server will respond with message about successfully set limit for all statistics, with a result set to 0 indicating success and an empty parameters field. If an error is encountered, the server returns a status code of 1 (error) and the text field contains the error description.

17.3.9. The statistic-sample-count-set Command

The statistic-sample-count-set command sets size based limit for collecting samples for given statistic. An example command may look like this:

{
    "command": "statistic-sample-count-set",
    "arguments": {
        "name": "pkt4-received",
        "max-samples": 100
    }

}

The server will respond with message about successfully set limit for the given statistic, with a result set to 0 indicating success and an empty parameters field. If an error is encountered (e.g. the requested statistic was not found), the server returns a status code of 1 (error) and the text field contains the error description.

17.3.10. The statistic-sample-count-set-all Command

The statistic-sample-count-set-all command sets size based limits for collecting samples for all statistics. An example command may look like this:

{
    "command": "statistic-sample-count-set-all",
    "arguments": {
        "max-samples": 100
    }

}

The server will respond with message about successfully set limit for all statistics, with a result set to 0 indicating success and an empty parameters field. If an error is encountered, the server returns a status code of 1 (error) and the text field contains the error description.

17.4. Time series

Previously, by default, each statistic held only a single data point. When Kea attempted to record a new value, the existing previous value was overwritten. That approach has the benefit of taking up little memory and it covers most cases reasonably well. However, there may be cases where you need to have many data points for some process. For example, some processes, such as received packet size, packet processing time or number of database queries needed to process a packet, are not cumulative and it would be useful to keep many data points, perhaps to do some form of statistical analysis afterwards.

Since Kea 1.6, by default, each statistic holds 20 data points. Setting such limit prevent unlimited memory consumption growth. There are two ways to define the limts: time based (e.g. keep samples from the last 5 minutes) and size based. It’s possible to change the size based limit by using one of two commands: statistic-sample-count-set, to set size limit for single statistic and statistic-sample-count-set-all for setting size based limits for all statistics. To set time based limit for single statistic use statistic-sample-age-set, and statistic-sample-age-set-all to set time based limits for all statistics. For given statistic only one type of limit can be active. It means that storage is limited only by time based limit or size based, never by both of them.