Storage backend (v2)¶

In order to implement a storage backend for cloudkitty, you’ll have to implement the following abstract class:

class cloudkitty.storage.v2.BaseStorage(*args, **kwargs)[source]¶

Abstract class for v2 storage objects.

static get_retention()[source]¶

Returns the retention period defined in the configuration.

Return type:	datetime.timedelta

init()[source]¶

Called for storage backend initialization

push(dataframes, scope_id=None)[source]¶

Pushes dataframes to the storage backend

A dataframe has the following format:

{
    "usage": {
        "bananas": [ # metric name
            {
                "vol": {
                    "unit": "banana",
                    "qty": 1
                },
                "rating": {
                    "price": 1
                },
                "groupby": {
                    "xxx_id": "hello",
                    "yyy_id": "bye",
                },
                "metadata": {
                    "flavor": "chocolate",
                    "eaten_by": "gorilla",
                },
           }
        ],
        "metric_name2": [...],
    }
   "period": {
        "begin": "1239781290", # timestamp
        "end": "1239793490", # timestamp
    }
}

Parameters:	dataframes (list) – List of dataframes

retrieve(begin=None, end=None, filters=None, group_filters=None, metric_types=None, offset=0, limit=1000, paginate=True)[source]¶

Returns the following dict:

{
   'total': int, # total amount of measures found
   'dataframes': list of dataframes,
}

Parameters:

begin (datetime) – Start date end (datetime) – End date filters (dict) – Metadata to filter on. ex: {‘flavor_id’: ‘42’} group_filters (dict) – Groupby to filter on. ex: {‘project_id’: ‘123ab’} metric_types (str or list) – Metric type to filter on. offset (int) – Offset for pagination limit (int) – Maximum amount of elements to return paginate (bool) – Defaults to True. If False, all found results will be returned.

Return type:

dict

total(groupby=None, begin=None, end=None, metric_types=None, filters=None, group_filters=None, offset=0, limit=1000, paginate=True)[source]¶

Returns a grouped total for given groupby.

Parameters:

groupby (list of strings) – Attributes on which to group by. These attributes must be part of the ‘groupby’ section for the given metric type in metrics.yml. In order to group by metric type, add ‘type’ to the groupby list. begin (datetime) – Start date end (datetime) – End date filters (dict) – Metadata to filter on. ex: {‘flavor_id’: ‘42’} group_filters (dict) – Groupby to filter on. ex: {‘project_id’: ‘123ab’} metric_types (str or list) – Metric type to filter on. offset (int) – Offset for pagination limit (int) – Maximum amount of elements to return paginate (bool) – Defaults to True. If False, all found results will be returned.

Return type:

dict

Returns a dict with the following format:

{
   'total': int, # total amount of results found
   'results': list of results,
}

Each result has the following format:

{
    'begin': XXX,
    'end': XXX,
    'rate': XXX,
    'groupby1': XXX,
    'groupby2': XXX
}

You’ll then need to register an entrypoint corresponding to your storage backend in the cloudkitty.storage.v2.backends section of the setup.cfg file.