Fix partial-frame-missing backfill logic

This had a bug prior where the end of a frame (a partial) wasn't being
sliced correctly and we'd get odd gaps showing up in the backfilled from
`brokerd` vs. tsdb end index. Repair this by doing timeframe aware index
diffing in `diff_history()` which seems to resolve it. Also, use the
frame-result's `end_dt: datetime` for the loop exit condition.
agg_feedz
Tyler Goodlet 2022-11-10 15:25:05 -05:00
parent 36868bb86e
commit 8476d8d056
1 changed files with 79 additions and 44 deletions

View File

@ -208,35 +208,46 @@ async def _setup_persistent_brokerd(
def diff_history( def diff_history(
array, array: np.ndarray,
start_dt, timeframe: int,
end_dt, start_dt: datetime,
last_tsdb_dt: Optional[datetime] = None end_dt: datetime,
last_tsdb_dt: datetime | None = None
) -> np.ndarray: ) -> np.ndarray:
to_push = array # no diffing with tsdb dt index possible..
if last_tsdb_dt is None:
return array
if last_tsdb_dt: time_attr = {
s_diff = (start_dt - last_tsdb_dt).seconds 1: 'seconds',
60: 'minutes',
}[timeframe]
i_diff = getattr((end_dt - last_tsdb_dt), time_attr)
# if we detect a partial frame's worth of data # if we detect a partial frame's worth of data
# that is new, slice out only that history and # that is new, slice out only that history and
# write to shm. # write to shm.
if ( if i_diff < 0:
s_diff < 0 # empty case since tsdb already has this history
): return array[:0]
if abs(s_diff) < len(array):
# the + 1 is because ``last_tsdb_dt`` is pulled from
# the last row entry for the ``'time'`` field retreived
# from the tsdb.
to_push = array[abs(s_diff) + 1:]
else: ln = len(array)
# pass back only the portion of the array that is to_push = array
# greater then the last time stamp in the tsdb.
time = array['time'] if i_diff < ln:
to_push = array[time >= last_tsdb_dt.timestamp()] # slice out missing history from end of frame
to_push = array[ln - i_diff:ln]
# XXX: OLD GAP HANDLING..if there's a gap in a partial frame
# worth. we don't need this any more with the timeframe aware
# diffing above right?
# else:
# # pass back only the portion of the array that is
# # greater then the last time stamp in the tsdb.
# time = array['time']
# to_push = array[time >= last_tsdb_dt.timestamp()]
log.debug( log.debug(
f'Pushing partial frame {to_push.size} to shm' f'Pushing partial frame {to_push.size} to shm'
@ -286,6 +297,7 @@ async def start_backfill(
to_push = diff_history( to_push = diff_history(
array, array,
timeframe,
start_dt, start_dt,
end_dt, end_dt,
last_tsdb_dt=last_tsdb_dt, last_tsdb_dt=last_tsdb_dt,
@ -334,12 +346,12 @@ async def start_backfill(
60: {'years': 6}, 60: {'years': 6},
} }
kwargs = periods[step_size_s] period_duration = periods[step_size_s]
# NOTE: manually set the "latest" datetime which we intend to # NOTE: manually set the "latest" datetime which we intend to
# backfill history "until" so as to adhere to the history # backfill history "until" so as to adhere to the history
# settings above when the tsdb is detected as being empty. # settings above when the tsdb is detected as being empty.
last_tsdb_dt = start_dt.subtract(**kwargs) last_tsdb_dt = start_dt.subtract(**period_duration)
# configure async query throttling # configure async query throttling
# rate = config.get('rate', 1) # rate = config.get('rate', 1)
@ -353,7 +365,7 @@ async def start_backfill(
# inline sequential loop where we simply pass the # inline sequential loop where we simply pass the
# last retrieved start dt to the next request as # last retrieved start dt to the next request as
# it's end dt. # it's end dt.
while start_dt > last_tsdb_dt: while end_dt > last_tsdb_dt:
log.debug( log.debug(
f'Requesting {step_size_s}s frame ending in {start_dt}' f'Requesting {step_size_s}s frame ending in {start_dt}'
) )
@ -363,6 +375,8 @@ async def start_backfill(
timeframe, timeframe,
end_dt=start_dt, end_dt=start_dt,
) )
# if timeframe == 1:
# await tractor.breakpoint()
# broker says there never was or is no more history to pull # broker says there never was or is no more history to pull
except DataUnavailable: except DataUnavailable:
@ -404,6 +418,7 @@ async def start_backfill(
to_push = diff_history( to_push = diff_history(
array, array,
timeframe,
start_dt, start_dt,
end_dt, end_dt,
last_tsdb_dt=last_tsdb_dt, last_tsdb_dt=last_tsdb_dt,
@ -424,6 +439,9 @@ async def start_backfill(
log.info( log.info(
f'Shm buffer overrun on: {start_dt} -> {end_dt}?' f'Shm buffer overrun on: {start_dt} -> {end_dt}?'
) )
# can't push the entire frame? so
# push only the amount that can fit..
await tractor.breakpoint()
break break
log.info( log.info(
@ -510,6 +528,8 @@ async def tsdb_backfill(
# start history anal and load missing new data via backend. # start history anal and load missing new data via backend.
for timeframe, shm in shms.items(): for timeframe, shm in shms.items():
# loads a (large) frame of data from the tsdb depending
# on the db's query size limit.
tsdb_history, first_tsdb_dt, last_tsdb_dt = await storage.load( tsdb_history, first_tsdb_dt, last_tsdb_dt = await storage.load(
fqsn, fqsn,
timeframe=timeframe, timeframe=timeframe,
@ -586,29 +606,45 @@ async def tsdb_backfill(
if bf_done: if bf_done:
await bf_done.wait() await bf_done.wait()
# Load tsdb history into shm buffer (for display).
# TODO: eventually it'd be nice to not require a shm array/buffer # TODO: eventually it'd be nice to not require a shm array/buffer
# to accomplish this.. maybe we can do some kind of tsdb direct to # to accomplish this.. maybe we can do some kind of tsdb direct to
# graphics format eventually in a child-actor? # graphics format eventually in a child-actor?
# do diff against last start frame of history and only fill
# in from the tsdb an allotment that allows for most recent
# to be loaded into mem *before* tsdb data.
if last_tsdb_dt and latest_start_dt:
dt_diff_s = (
latest_start_dt - last_tsdb_dt
).seconds
else:
dt_diff_s = 0
# TODO: see if there's faster multi-field reads: # TODO: see if there's faster multi-field reads:
# https://numpy.org/doc/stable/user/basics.rec.html#accessing-multiple-fields # https://numpy.org/doc/stable/user/basics.rec.html#accessing-multiple-fields
# re-index with a `time` and index field # re-index with a `time` and index field
prepend_start = shm._first.value prepend_start = shm._first.value
shm_last_dt = pendulum.from_timestamp(shm.array[0]['time'])
# sanity check on most-recent-data loading if last_tsdb_dt:
assert prepend_start > dt_diff_s assert shm_last_dt >= last_tsdb_dt
# do diff against start index of last frame of history and only
# fill in an amount of datums from tsdb allows for most recent
# to be loaded into mem *before* tsdb data.
if (
last_tsdb_dt
and latest_start_dt
):
backfilled_size_s = (
latest_start_dt - last_tsdb_dt
).seconds
else:
backfilled_size_s = (
latest_start_dt - shm_last_dt
).seconds
# Load TSDB history into shm buffer (for display) if there is
# remaining buffer space.
# if the shm buffer len is not large enough to contain
# all missing data between the most recent backend-queried frame
# and the most recent dt-index in the db we warn that we only
# want to load a portion of the next tsdb query to fill that
# space.
log.info(
f'{backfilled_size_s} seconds worth of {timeframe}s data loaded'
)
if ( if (
len(tsdb_history) len(tsdb_history)
@ -1266,7 +1302,6 @@ async def open_feed_bus(
# expected to append it's own name to the fqsn, so we filter # expected to append it's own name to the fqsn, so we filter
# on keys which *do not* include that name (e.g .ib) . # on keys which *do not* include that name (e.g .ib) .
bus._subscribers.setdefault(bfqsn, []) bus._subscribers.setdefault(bfqsn, [])
# await tractor.breakpoint()
# sync feed subscribers with flume handles # sync feed subscribers with flume handles
await ctx.started( await ctx.started(