README

A small library that helps store graph-type (incremental stats) in mongo database using doctrine ODM

Installation

Add this to your composer.json

"require": {
    "bigtallbill/php-mongo-graph-model": "dev-master"
}

Usage

// instantiate a new instance providing the doctrine DocumentManager instance
$gm = new GraphManager($dm);

$gm->setGranularity(GraphManager::MINUTE);
$gm->setWindow(GraphManager::HOUR);

// create a random statistic to store. These are instances of ParsedStat
// and can be created manually if required
$stat = $gm->parseSimpleKeyValue('system_stats', 'cpu', rand(25, 50));

// finally run an upsert using doctrine to create or update the stat
$gm->insertStat($stat);

How this model looks

Using a granularity of Hours and a window of Days:

{
    "_id" : ObjectId("5560b3b7919124c4e03710fd"),
    "date" : ISODate("2015-05-23T00:00:00.000Z"),
    "granularity" : NumberLong(3600),
    "group" : "system_stats",
    "key" : "mem",
    "window" : NumberLong(86400),
    "window_human" : "Day",
    "granularity_human" : "Hour",
    "segments" : {
        "1432400400" : NumberLong(108588),
        "1432404000" : NumberLong(119624)
    }
}

Under the segments key is where each segment of granularity is stored (in our case the top of the current hour). The date is locked to the top of the current window (in our case the current day)

Here is another example with a granularity of minutes and a window of hours:

{
    "_id" : ObjectId("5560c023919124c4e0371102"),
    "date" : ISODate("2015-05-23T18:00:00.000Z"),
    "granularity" : NumberLong(60),
    "group" : "system_stats",
    "key" : "processes",
    "window" : NumberLong(3600),
    "window_human" : "Hour",
    "granularity_human" : "Minute",
    "segments" : {
        "1432404000" : NumberLong(59041),
        "1432404060" : NumberLong(41138),
        "1432404120" : NumberLong(62544),
        "1432404180" : NumberLong(59392),
        "1432404240" : NumberLong(22948),
        "1432404300" : NumberLong(58825),
        "1432404360" : NumberLong(31936),
        "1432404420" : NumberLong(17407),
        "1432404480" : NumberLong(30097),
        "1432404540" : NumberLong(54094),
        "1432404600" : NumberLong(34739),
        "1432404660" : NumberLong(34164),
        "1432404720" : NumberLong(34571),
        "1432404780" : NumberLong(28592)
    }
}