Get multiple counters

Home - HTTP Interface - Get multiple counters

Usage

POST /*

Parameters

The POST data contains a JSON object value, sort of like this:

{
    "start": START,
    "end": END,
    "maxSamples": MAX,
    "keys": ["counter.name", "yet.another.counter.name"],
    "trailing": 0
}

These are the entries that can appear in the object:

• keys - A required list of counter names to grab data for. This is an array of strings.
• maxSamples - The maximum number of samples that should be returned by this search. Used to simplify datasets for graphing.
• start - An optional Unix timestamp specifying the lower time bound to use. (Default: 15 minutes before end)
• end - An optional Unix timestamp specifying the upper time bound to use. (Default: now)
• trailing - An optional value of 0 or 1 to read trailing averages (seasonal, lambda-based values) instead of regular counter history.
• The value of end must be >= start.

Overview

Use this URI to get data for multiple counters at once, and then down-size the number of samples. It's extremely useful for generating fixed-size graphs with multiple plots.

Returns the data for each counter specified, in succession.

Response

When the request is well-formed, with all of its required parameters,

• Status code: 200

Response body:

{
    // A list of all names that match the pattern given.
    "start": int, // The start time that was passed to this request, or 0 if none was given.
    "end":   int, // The end time that was passed to this request, or 0 if none was given.
    // The name of a counter that was asked for in the 'keys' parameter.
    "counter.name": {
        "interval": int,
        // A list of buckets of data within this time range. Possibly empty.
        "data": [
            // One bucket of data in the list. 
            {
                "time":  int,   // When this bucket was recorded. A unix timestamp.
                "count": int,   // Number of values in the bucket.
                "min":   float, // Minimum of all values in the bucket.
                "max":   float, // Maximum of all values in the bucket.
                "sum":   float, // Sum of each value in this bucket.
                "sumsq": float, // Sum of each value squared in this bucket.
                "avg":   float, // The average of the bucket, or the number 0 if the count is zero.
                "sdev":  float  // The standard deviation of the bucket, or the number 0 if the count is less than 2.
            }
            // ... the rest of the data. There will be at most maxSamples number of data entries for this counter.
        ]
    },
    // This is what you get if a counter does not exist.
    "something.that.does.not.exist": "Not found",
    // ... the rest of the keys
}

When a request that does not specify `keys` in its POST data,

• Status code: 400

Trailing counters

If you have configured seasonal averages, the trailing=1 parameter will request those instead of the regular counter history. If there is more than one season, the shortest season that covers the given time interval will be selected.