diff --git a/cloudkitty/storage/v2/elasticsearch/__init__.py b/cloudkitty/storage/v2/elasticsearch/__init__.py
index 709457d2..7b28acff 100644
--- a/cloudkitty/storage/v2/elasticsearch/__init__.py
+++ b/cloudkitty/storage/v2/elasticsearch/__init__.py
@@ -184,6 +184,9 @@ class ElasticsearchStorage(v2_storage.BaseStorage):
         # Means we had a composite aggregation
         if 'key' in doc.keys():
             for key, value in doc['key'].items():
+                if key == 'begin' or key == 'end':
+                    # Elasticsearch returns ts in milliseconds
+                    value = tzutils.dt_from_ts(value // 1000)
                 output[key] = value
         return output
 
diff --git a/cloudkitty/storage/v2/elasticsearch/client.py b/cloudkitty/storage/v2/elasticsearch/client.py
index 3a1aa8dc..1fff1cf2 100644
--- a/cloudkitty/storage/v2/elasticsearch/client.py
+++ b/cloudkitty/storage/v2/elasticsearch/client.py
@@ -92,14 +92,18 @@ class ElasticsearchClient(object):
                            {'term': {'metadata.' + k: v}}]
         return should
 
-    @staticmethod
-    def _build_composite(groupby):
+    def _build_composite(self, groupby):
         if not groupby:
             return []
         sources = []
         for elem in groupby:
             if elem == 'type':
                 sources.append({'type': {'terms': {'field': 'type'}}})
+            elif elem == 'time':
+                # Not doing a date_histogram aggregation because we don't know
+                # the period
+                sources.append({'begin': {'terms': {'field': 'start'}}})
+                sources.append({'end': {'terms': {'field': 'end'}}})
             else:
                 sources.append({elem: {'terms': {'field': 'groupby.' + elem}}})
 
diff --git a/cloudkitty/storage/v2/influx.py b/cloudkitty/storage/v2/influx.py
index 602dc1e8..7cab8293 100644
--- a/cloudkitty/storage/v2/influx.py
+++ b/cloudkitty/storage/v2/influx.py
@@ -65,19 +65,27 @@ CONF.register_opts(influx_storage_opts, INFLUX_STORAGE_GROUP)
 PERIOD_FIELD_NAME = '__ck_collect_period'
 
 
+def _sanitized_groupby(groupby):
+    forbidden = ('time',)
+    return [g for g in groupby if g not in forbidden] if groupby else []
+
+
 class InfluxClient(object):
     """Classe used to ease interaction with InfluxDB"""
 
-    def __init__(self, chunk_size=500, autocommit=True):
+    def __init__(self, chunk_size=500, autocommit=True, default_period=3600):
         """Creates an InfluxClient object.
 
         :param chunk_size: Size after which points should be pushed.
         :param autocommit: Set to false to disable autocommit
+        :param default_period: Placeholder for the period in cae it can't
+                               be determined.
         """
         self._conn = self._get_influx_client()
         self._chunk_size = chunk_size
         self._autocommit = autocommit
         self._retention_policy = CONF.storage_influxdb.retention_policy
+        self._default_period = default_period
         self._points = []
 
     @staticmethod
@@ -185,7 +193,13 @@ class InfluxClient(object):
         query += self._get_type_query(types)
 
         if groupby:
-            groupby_query = '"' + '","'.join(groupby) + '"'
+            groupby_query = ''
+            if 'time' in groupby:
+                groupby_query += 'time(' + str(self._default_period) + 's)'
+                groupby_query += ',' if groupby else ''
+            if groupby:
+                groupby_query += '"' + '","'.join(
+                    _sanitized_groupby(groupby)) + '"'
             query += ' GROUP BY ' + groupby_query
 
         query += ';'
@@ -243,7 +257,7 @@ class InfluxStorage(v2_storage.BaseStorage):
     def __init__(self, *args, **kwargs):
         super(InfluxStorage, self).__init__(*args, **kwargs)
         self._default_period = kwargs.get('period') or CONF.collect.period
-        self._conn = InfluxClient()
+        self._conn = InfluxClient(default_period=self._default_period)
 
     def init(self):
         policy = CONF.storage_influxdb.retention_policy
@@ -326,8 +340,11 @@ class InfluxStorage(v2_storage.BaseStorage):
     def delete(self, begin=None, end=None, filters=None):
         self._conn.delete(begin, end, filters)
 
-    @staticmethod
-    def _get_total_elem(begin, end, groupby, series_groupby, point):
+    def _get_total_elem(self, begin, end, groupby, series_groupby, point):
+        if groupby and 'time' in groupby:
+            begin = tzutils.dt_from_iso(point['time'])
+            period = point.get(PERIOD_FIELD_NAME) or self._default_period
+            end = tzutils.add_delta(begin, datetime.timedelta(seconds=period))
         output = {
             'begin': begin,
             'end': end,
@@ -335,7 +352,7 @@ class InfluxStorage(v2_storage.BaseStorage):
             'rate': point['price'],
         }
         if groupby:
-            for group in groupby:
+            for group in _sanitized_groupby(groupby):
                 output[group] = series_groupby.get(group, '')
         return output
 
@@ -353,12 +370,18 @@ class InfluxStorage(v2_storage.BaseStorage):
         output = []
         for (series_name, series_groupby), points in total.items():
             for point in points:
-                output.append(self._get_total_elem(
-                    begin, end,
-                    groupby,
-                    series_groupby,
-                    point))
+                # NOTE(peschk_l): InfluxDB returns all timestamps for a given
+                # period and interval, even those with no data. This filters
+                # out periods with no data
+                if point['qty'] is not None and point['price'] is not None:
+                    output.append(self._get_total_elem(
+                        tzutils.utc_to_local(begin),
+                        tzutils.utc_to_local(end),
+                        groupby,
+                        series_groupby,
+                        point))
 
+        groupby = _sanitized_groupby(groupby)
         if groupby:
             output.sort(key=lambda x: [x[group] for group in groupby])
         return {
diff --git a/cloudkitty/tests/gabbi/gabbits/v2-summary.yaml b/cloudkitty/tests/gabbi/gabbits/v2-summary.yaml
index 5005d58b..b584fcef 100644
--- a/cloudkitty/tests/gabbi/gabbits/v2-summary.yaml
+++ b/cloudkitty/tests/gabbi/gabbits/v2-summary.yaml
@@ -68,3 +68,21 @@ tests:
     response_json_paths:
       $.results.`len`: 0
       $.total: 0
+
+  - name: Get a summary grouped by time
+    url: /v2/summary
+    status: 200
+    query_parameters:
+      groupby: [time]
+    response_json_paths:
+      $.results.`len`: 1
+      $.total: 1
+
+  - name: Get a summary grouped by time and project_id
+    url: /v2/summary
+    status: 200
+    query_parameters:
+      groupby: [time, project_id]
+    response_json_paths:
+      $.results.`len`: 2
+      $.total: 2
diff --git a/cloudkitty/tests/storage/v2/influx_utils.py b/cloudkitty/tests/storage/v2/influx_utils.py
index d7829289..de99c7ec 100644
--- a/cloudkitty/tests/storage/v2/influx_utils.py
+++ b/cloudkitty/tests/storage/v2/influx_utils.py
@@ -17,6 +17,7 @@ import functools
 
 from influxdb import resultset
 
+from cloudkitty.storage.v2.influx import _sanitized_groupby
 from cloudkitty.storage.v2.influx import InfluxClient
 
 
@@ -63,6 +64,12 @@ class FakeInfluxClient(InfluxClient):
                 break
             valid = True
             for tag in serie['tags'].keys():
+                if tag == 'time':
+                    if point['time'].isoformat() != serie['values'][0][0]:
+                        valid = False
+                        break
+                    else:
+                        continue
                 if tag not in point['tags'].keys() or \
                    point['tags'][tag] != serie['tags'][tag]:
                     valid = False
@@ -74,10 +81,11 @@ class FakeInfluxClient(InfluxClient):
         if target_serie is None:
             target_serie = copy.deepcopy(self.total_series_sample)
             if groupby:
-                target_serie['tags'] = {k: point['tags'][k] for k in groupby}
+                target_serie['tags'] = {k: point['tags'][k] for k in
+                                        _sanitized_groupby(groupby)}
             else:
                 target_serie['tags'] = {}
-            target_serie['values'] = [['1970-01-01T00:00:00Z', 0, 0]]
+            target_serie['values'] = [[point['time'].isoformat(), 0, 0]]
             series.append(target_serie)
         return target_serie
 
diff --git a/cloudkitty/tests/test_tzutils.py b/cloudkitty/tests/test_tzutils.py
index 2ebc1362..5ebb77db 100644
--- a/cloudkitty/tests/test_tzutils.py
+++ b/cloudkitty/tests/test_tzutils.py
@@ -16,6 +16,7 @@ import datetime
 import unittest
 
 from dateutil import tz
+import mock
 from oslo_utils import timeutils
 
 from cloudkitty import tzutils
@@ -132,3 +133,15 @@ class TestTZUtils(unittest.TestCase):
         two = datetime.datetime(2019, 3, 31, 3,
                                 tzinfo=tz.gettz('Europe/Paris'))
         self.assertEqual(tzutils.diff_seconds(two, one), 3600)
+
+    def test_cloudkitty_dt_from_ts_as_utc(self):
+        ts = 1569902400
+        dt = datetime.datetime(2019, 10, 1, 4, tzinfo=tz.UTC)
+        self.assertEqual(dt, tzutils.dt_from_ts(ts, as_utc=True))
+
+    def test_cloudkitty_dt_from_ts_local_tz(self):
+        ts = 1569902400
+        timezone = tz.gettz('Europe/Paris')
+        dt = datetime.datetime(2019, 10, 1, 6, tzinfo=timezone)
+        with mock.patch.object(tzutils, '_LOCAL_TZ', new=timezone):
+            self.assertEqual(dt, tzutils.dt_from_ts(ts))
diff --git a/cloudkitty/tzutils.py b/cloudkitty/tzutils.py
index 6f143e64..54392280 100644
--- a/cloudkitty/tzutils.py
+++ b/cloudkitty/tzutils.py
@@ -84,6 +84,14 @@ def dt_from_iso(time_str, as_utc=False):
         tz.UTC if as_utc else _LOCAL_TZ).replace(microsecond=0)
 
 
+def dt_from_ts(ts, as_utc=False):
+    """Parses a timezone-aware datetime object from an epoch timestamp.
+
+    Returns the object as being from the local timezone.
+    """
+    return datetime.datetime.fromtimestamp(ts, tz.UTC if as_utc else _LOCAL_TZ)
+
+
 def add_delta(dt, delta):
     """Adds a timedelta to a datetime object.
 
diff --git a/doc/source/api-reference/v2/api_samples/summary/summary_get_groupby_time.json b/doc/source/api-reference/v2/api_samples/summary/summary_get_groupby_time.json
new file mode 100644
index 00000000..85dff158
--- /dev/null
+++ b/doc/source/api-reference/v2/api_samples/summary/summary_get_groupby_time.json
@@ -0,0 +1,33 @@
+{
+  "total": 232,
+  "columns": [
+    "begin",
+    "end",
+    "qty",
+    "rate",
+    "project_id"
+  ],
+  "results": [
+    [
+      "2019-10-01T06:00:00+02:00",
+      "2019-10-01T07:00:00+02:00",
+      3.5533905029296875,
+      1.7766952514648438,
+      "84631866b2d84db49b29828052bdc287"
+    ],
+    [
+      "2019-10-01T07:00:00+02:00",
+      "2019-10-01T08:00:00+02:00",
+      3.5533905029296875,
+      1.7766952514648438,
+      "84631866b2d84db49b29828052bdc287"
+    ],
+    [
+      "2019-10-01T08:00:00+02:00",
+      "2019-10-01T09:00:00+02:00",
+      3.5533905029296875,
+      1.7766952514648438,
+      "84631866b2d84db49b29828052bdc287"
+    ]
+  ]
+}
diff --git a/doc/source/api-reference/v2/summary/summary.inc b/doc/source/api-reference/v2/summary/summary.inc
index 3d897672..99ed7fde 100644
--- a/doc/source/api-reference/v2/summary/summary.inc
+++ b/doc/source/api-reference/v2/summary/summary.inc
@@ -65,6 +65,11 @@ the columns for each element of ``results``. The columns are the four mandatory
 (``begin``, ``end``, ``qty``, ``rate``) along with each attribute the result is
 grouped by.
 
+.. note:: It is also possible to group data by time, in order to obtain timeseries.
+          In order to do this, group by ``time``. No extra column will be added,
+          but you'll get one entry per collect period in the queried timeframe.
+          See examples below.
+
 .. rest_parameters:: summary/summary_parameters.yml
 
    - begin: begin_resp
@@ -75,6 +80,15 @@ grouped by.
 Response Example
 ----------------
 
+Grouping by time and project_id:
+
+.. code-block:: shell
+
+   curl "http://cloudkitty-api:8889/v2/summary?groupby=time&groupby=project_id&limit=3"
+
+.. literalinclude:: ./api_samples/summary/summary_get_groupby_time.json
+   :language: javascript
+
 .. code-block:: shell
 
    curl "http://cloudkitty-api:8889/v2/summary?filters=project_id%3Afe9c35372db6420089883805b37a34af&groupby=type&groupby=project_id"
diff --git a/releasenotes/notes/support-groupby-time-v2-summary-48ff5ad671f8c7c5.yaml b/releasenotes/notes/support-groupby-time-v2-summary-48ff5ad671f8c7c5.yaml
new file mode 100644
index 00000000..aa9b4897
--- /dev/null
+++ b/releasenotes/notes/support-groupby-time-v2-summary-48ff5ad671f8c7c5.yaml
@@ -0,0 +1,6 @@
+---
+upgrade:
+  - |
+    It is now possible to group v2 summaries by timestamp. In order to do this,
+    the ``time`` parameter must be specified in the ``groupby`` list:
+    ``cloudkitty summary get -g time,type``.