pikers
/
piker
3
0
Fork 0
Code Issues 2 Pull Requests 4 Packages Projects Releases Wiki Activity

Make `fsp.cascade()` expect src/dst `Flume`s 

Been meaning to this for a while, and there's still a few design
/ interface kinks (like `.mkt: MktPair` which should be better
generalized?) but this flips over all of the fsp chaining engine
to operate on the higher level `Flume` APIs via the newly cobbled
`Cascade` thinger..
distribute_dis
Tyler Goodlet 2023-12-06 17:53:35 -05:00
parent 9e71e0768f
commit dd0167b9a5
3 changed files with 116 additions and 93 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
piker/fsp/__init__.py
View File

@ -26,7 +26,10 @@ from ._api import (
maybe_mk_fsp_shm,
Fsp,
)
from ._engine import cascade
from ._engine import (
cascade,
Cascade,
)
from ._volume import (
dolla_vlm,
flow_rates,
@ -35,6 +38,7 @@ from ._volume import (
__all__: list[str] = [
'cascade',
'Cascade',
'maybe_mk_fsp_shm',
'Fsp',
'dolla_vlm',
@ -46,9 +50,12 @@ __all__: list[str] = [
async def latency(
source: 'TickStream[Dict[str, float]]', # noqa
ohlcv: np.ndarray
) -> AsyncIterator[np.ndarray]:
"""Latency measurements, broker to piker.
"""
'''
Latency measurements, broker to piker.
'''
# TODO: do we want to offer yielding this async
# before the rt data connection comes up?

172
piker/fsp/_engine.py
View File

@ -24,8 +24,6 @@ from functools import partial
from typing import (
AsyncIterator,
Callable,
Optional,
Union,
)
import numpy as np
@ -37,7 +35,6 @@ from tractor.msg import NamespacePath
from piker.types import Struct
from ..log import get_logger, get_console_log
from .. import data
from ..data import attach_shm_array
from ..data.feed import (
Flume,
Feed,
@ -117,8 +114,8 @@ class Cascade(Struct):
'''
# TODO: make these `Flume`s
src: ShmArray
dst: ShmArray
src: Flume
dst: Flume
tn: trio.Nursery
fsp: Fsp # UI-side middleware ctl API
@ -139,11 +136,12 @@ class Cascade(Struct):
# always trigger UI refresh after history update,
# see ``piker.ui._fsp.FspAdmin.open_chain()`` and
# ``piker.ui._display.trigger_update()``.
dst_shm: ShmArray = self.dst.rt_shm
await self.client_stream.send({
'fsp_update': {
'key': self.dst.token,
'first': self.dst._first.value,
'last': self.dst._last.value,
'key': dst_shm.token,
'first': dst_shm._first.value,
'last': dst_shm._last.value,
}
})
return index
@ -154,10 +152,10 @@ class Cascade(Struct):
output array is aligned to its source array.
'''
src: ShmArray = self.src
dst: ShmArray = self.dst
step_diff = src.index - dst.index
len_diff = abs(len(src.array) - len(dst.array))
src_shm: ShmArray = self.src.rt_shm
dst_shm: ShmArray = self.dst.rt_shm
step_diff = src_shm.index - dst_shm.index
len_diff = abs(len(src_shm.array) - len(dst_shm.array))
synced: bool = not (
# the source is likely backfilling and we must
# sync history calculations
@ -172,7 +170,7 @@ class Cascade(Struct):
fsp: Fsp = self.fsp
log.warning(
'***DESYNCED FSP***\n'
f'{fsp.ns_path}@{src.token}\n'
f'{fsp.ns_path}@{src_shm.token}\n'
f'step_diff: {step_diff}\n'
f'len_diff: {len_diff}\n'
)
@ -183,10 +181,10 @@ class Cascade(Struct):
)
async def poll_and_sync_to_step(self) -> int:
synced, step_diff, _ = self.is_synced() #src, dst)
synced, step_diff, _ = self.is_synced()
while not synced:
await self.resync()
synced, step_diff, _ = self.is_synced() #src, dst)
synced, step_diff, _ = self.is_synced()
return step_diff
@ -203,16 +201,13 @@ class Cascade(Struct):
async def connect_streams(
casc: Cascade,
mkt: MktPair,
flume: Flume,
quote_stream: trio.abc.ReceiveChannel,
src: Flume,
dst: Flume,
src: ShmArray,
dst: ShmArray,
func: Callable,
edge_func: Callable,
# attach_stream: bool = False,
task_status: TaskStatus[None] = trio.TASK_STATUS_IGNORED,
@ -226,7 +221,7 @@ async def connect_streams(
Not literally, but something like:
func(Flume_in) -> Flume_out
edge_func(Flume_in) -> Flume_out
'''
profiler = Profiler(
@ -234,12 +229,14 @@ async def connect_streams(
disabled=True
)
fqme: str = mkt.fqme
# TODO: just pull it from src.mkt.fqme no?
# fqme: str = mkt.fqme
fqme: str = src.mkt.fqme
# TODO: dynamic introspection of what the underlying (vertex)
# function actually requires from input node (flumes) then
# deliver those inputs as part of a graph "compilation" step?
out_stream = func(
out_stream = edge_func(
# TODO: do we even need this if we do the feed api right?
# shouldn't a local stream do this before we get a handle
@ -249,19 +246,19 @@ async def connect_streams(
# XXX: currently the ``ohlcv`` arg, but we should allow
# (dynamic) requests for src flume (node) streams?
flume.rt_shm,
src.rt_shm,
)
# HISTORY COMPUTE PHASE
# conduct a single iteration of fsp with historical bars input
# and get historical output.
history_output: Union[
dict[str, np.ndarray], # multi-output case
np.ndarray, # single output case
]
history_output: (
dict[str, np.ndarray] # multi-output case
| np.ndarray, # single output case
)
history_output = await anext(out_stream)
func_name = func.__name__
func_name = edge_func.__name__
profiler(f'{func_name} generated history')
# build struct array with an 'index' field to push as history
@ -269,10 +266,12 @@ async def connect_streams(
# TODO: push using a[['f0', 'f1', .., 'fn']] = .. syntax no?
# if the output array is multi-field then push
# each respective field.
fields = getattr(dst.array.dtype, 'fields', None).copy()
dst_shm: ShmArray = dst.rt_shm
fields = getattr(dst_shm.array.dtype, 'fields', None).copy()
fields.pop('index')
history_by_field: Optional[np.ndarray] = None
src_time = src.array['time']
history_by_field: np.ndarray | None = None
src_shm: ShmArray = src.rt_shm
src_time = src_shm.array['time']
if (
fields and
@ -291,7 +290,7 @@ async def connect_streams(
if history_by_field is None:
if output is None:
length = len(src.array)
length = len(src_shm.array)
else:
length = len(output)
@ -300,7 +299,7 @@ async def connect_streams(
# will be pushed to shm.
history_by_field = np.zeros(
length,
dtype=dst.array.dtype
dtype=dst_shm.array.dtype
)
if output is None:
@ -317,13 +316,13 @@ async def connect_streams(
)
history_by_field = np.zeros(
len(history_output),
dtype=dst.array.dtype
dtype=dst_shm.array.dtype
)
history_by_field[func_name] = history_output
history_by_field['time'] = src_time[-len(history_by_field):]
history_output['time'] = src.array['time']
history_output['time'] = src_shm.array['time']
# TODO: XXX:
# THERE'S A BIG BUG HERE WITH THE `index` field since we're
@ -336,11 +335,11 @@ async def connect_streams(
# is `index` aware such that historical data can be indexed
# relative to the true first datum? Not sure if this is sane
# for incremental compuations.
first = dst._first.value = src._first.value
first = dst_shm._first.value = src_shm._first.value
# TODO: can we use this `start` flag instead of the manual
# setting above?
index = dst.push(
index = dst_shm.push(
history_by_field,
start=first,
)
@ -367,12 +366,12 @@ async def connect_streams(
log.debug(f"{func_name}: {processed}")
key, output = processed
# dst.array[-1][key] = output
dst.array[[key, 'time']][-1] = (
dst_shm.array[[key, 'time']][-1] = (
output,
# TODO: what about pushing ``time.time_ns()``
# in which case we'll need to round at the graphics
# processing / sampling layer?
src.array[-1]['time']
src_shm.array[-1]['time']
)
# NOTE: for now we aren't streaming this to the consumer
@ -384,7 +383,7 @@ async def connect_streams(
# N-consumers who subscribe for the real-time output,
# which we'll likely want to implement using local-mem
# chans for the fan out?
# index = src.index
# index = src_shm.index
# if attach_stream:
# await client_stream.send(index)
@ -405,16 +404,15 @@ async def cascade(
# data feed key
fqme: str,
# TODO: expect and attach from `Flume.to_msg()`s!
src_shm_token: dict,
dst_shm_token: tuple[str, np.dtype],
# flume pair cascaded using an "edge function"
src_flume_addr: dict,
dst_flume_addr: dict,
ns_path: NamespacePath,
shm_registry: dict[str, _Token],
zero_on_step: bool = False,
loglevel: Optional[str] = None,
loglevel: str | None = None,
) -> None:
'''
@ -430,8 +428,14 @@ async def cascade(
if loglevel:
get_console_log(loglevel)
src: ShmArray = attach_shm_array(token=src_shm_token)
dst: ShmArray = attach_shm_array(readonly=False, token=dst_shm_token)
src: Flume = Flume.from_msg(src_flume_addr)
dst: Flume = Flume.from_msg(
dst_flume_addr,
readonly=False,
)
# src: ShmArray = attach_shm_array(token=src_shm_token)
# dst: ShmArray = attach_shm_array(readonly=False, token=dst_shm_token)
reg = _load_builtins()
lines = '\n'.join([f'{key.rpartition(":")[2]} => {key}' for key in reg])
@ -439,11 +443,11 @@ async def cascade(
f'Registered FSP set:\n{lines}'
)
# update actorlocal flows table which registers
# readonly "instances" of this fsp for symbol/source
# so that consumer fsps can look it up by source + fsp.
# TODO: ugh i hate this wind/unwind to list over the wire
# but not sure how else to do it.
# NOTE XXX: update actorlocal flows table which registers
# readonly "instances" of this fsp for symbol/source so that
# consumer fsps can look it up by source + fsp.
# TODO: ugh i hate this wind/unwind to list over the wire but
# not sure how else to do it.
for (token, fsp_name, dst_token) in shm_registry:
Fsp._flow_registry[(
_Token.from_msg(token),
@ -459,6 +463,9 @@ async def cascade(
# TODO: assume it's a func target path
raise ValueError(f'Unknown fsp target: {ns_path}')
_fqme: str = src.mkt.fqme
assert _fqme == fqme
# open a data feed stream with requested broker
feed: Feed
async with data.feed.maybe_open_feed(
@ -472,12 +479,21 @@ async def cascade(
) as feed:
flume = feed.flumes[fqme]
mkt = flume.mkt
flume: Flume = feed.flumes[fqme]
# XXX: can't do this since flume.feed will be set XD
# assert flume == src
assert flume.mkt == src.mkt
mkt: MktPair = flume.mkt
# NOTE: FOR NOW, sanity checks around the feed as being
# always the src flume (until we get to fancier/lengthier
# chains/graphs.
assert src.rt_shm.token == flume.rt_shm.token
# XXX: won't work bc the _hist_shm_token value will be
# list[list] after IPC..
# assert flume.to_msg() == src_flume_addr
# TODO: make an equivalent `Flume` around the Fsp output
# streams and chain them using a `Cascade` Bo
assert src.token == flume.rt_shm.token
profiler(f'{func}: feed up')
func_name: str = func.__name__
@ -497,34 +513,34 @@ async def cascade(
# TODO: this seems like it should be wrapped somewhere?
fsp_target = partial(
connect_streams,
casc=casc,
mkt=mkt,
flume=flume,
quote_stream=flume.stream,
# shm
# flumes and shm passthrough
src=src,
dst=dst,
# chain function which takes src flume input(s)
# and renders dst flume output(s)
func=func
edge_func=func
)
async with casc.open_edge(
bind_func=fsp_target,
) as index:
# casc.bind_func = fsp_target
# index = await tn.start(fsp_target)
dst_shm: ShmArray = dst.rt_shm
src_shm: ShmArray = src.rt_shm
if zero_on_step:
last = dst.array[-1:]
last = dst.rt_shm.array[-1:]
zeroed = np.zeros(last.shape, dtype=last.dtype)
profiler(f'{func_name}: fsp up')
# sync client
# sync to client-side actor
await ctx.started(index)
# XXX: rt stream with client which we MUST
@ -532,24 +548,26 @@ async def cascade(
# incremental "updates" as history prepends take
# place.
async with ctx.open_stream() as client_stream:
casc.client_stream = client_stream
casc.client_stream: tractor.MsgStream = client_stream
s, step, ld = casc.is_synced() #src, dst)
s, step, ld = casc.is_synced()
# detect sample period step for subscription to increment
# signal
times = src.array['time']
times = src.rt_shm.array['time']
if len(times) > 1:
last_ts = times[-1]
delay_s = float(last_ts - times[times != last_ts][-1])
delay_s: float = float(last_ts - times[times != last_ts][-1])
else:
# our default "HFT" sample rate.
delay_s = _default_delay_s
delay_s: float = _default_delay_s
# sub and increment the underlying shared memory buffer
# on every step msg received from the global `samplerd`
# service.
async with open_sample_stream(float(delay_s)) as istream:
async with open_sample_stream(
float(delay_s)
) as istream:
profiler(f'{func_name}: sample stream up')
profiler.finish()
@ -560,7 +578,7 @@ async def cascade(
# respawn the compute task if the source
# array has been updated such that we compute
# new history from the (prepended) source.
synced, step_diff, _ = casc.is_synced() #src, dst)
synced, step_diff, _ = casc.is_synced()
if not synced:
step_diff: int = await casc.poll_and_sync_to_step()
@ -570,7 +588,7 @@ async def cascade(
continue
# read out last shm row, copy and write new row
array = dst.array
array = dst_shm.array
# some metrics like vlm should be reset
# to zero every step.
@ -579,14 +597,14 @@ async def cascade(
else:
last = array[-1:].copy()
dst.push(last)
dst.rt_shm.push(last)
# sync with source buffer's time step
src_l2 = src.array[-2:]
src_l2 = src_shm.array[-2:]
src_li, src_lt = src_l2[-1][['index', 'time']]
src_2li, src_2lt = src_l2[-2][['index', 'time']]
dst._array['time'][src_li] = src_lt
dst._array['time'][src_2li] = src_2lt
dst_shm._array['time'][src_li] = src_lt
dst_shm._array['time'][src_2li] = src_2lt
# last2 = dst.array[-2:]
# if (

24
piker/ui/_fsp.py
View File

@ -390,7 +390,7 @@ class FspAdmin:
complete: trio.Event,
started: trio.Event,
fqme: str,
dst_fsp_flume: Flume,
dst_flume: Flume,
conf: dict,
target: Fsp,
loglevel: str,
@ -408,16 +408,14 @@ class FspAdmin:
# chaining entrypoint
cascade,
# TODO: can't we just drop this and expect
# far end to read the src flume's .mkt.fqme?
# data feed key
fqme=fqme,
# TODO: pass `Flume.to_msg()`s here?
# mems
src_shm_token=self.flume.rt_shm.token,
dst_shm_token=dst_fsp_flume.rt_shm.token,
# target
ns_path=ns_path,
src_flume_addr=self.flume.to_msg(),
dst_flume_addr=dst_flume.to_msg(),
ns_path=ns_path, # edge-bind-func
loglevel=loglevel,
zero_on_step=conf.get('zero_on_step', False),
@ -431,14 +429,14 @@ class FspAdmin:
ctx.open_stream() as stream,
):
dst_fsp_flume.stream: tractor.MsgStream = stream
dst_flume.stream: tractor.MsgStream = stream
# register output data
self._registry[
(fqme, ns_path)
] = (
stream,
dst_fsp_flume.rt_shm,
dst_flume.rt_shm,
complete
)
@ -515,7 +513,7 @@ class FspAdmin:
broker='piker',
_atype='fsp',
)
dst_fsp_flume = Flume(
dst_flume = Flume(
mkt=mkt,
_rt_shm_token=dst_shm.token,
first_quote={},
@ -543,13 +541,13 @@ class FspAdmin:
complete,
started,
fqme,
dst_fsp_flume,
dst_flume,
conf,
target,
loglevel,
)
return dst_fsp_flume, started
return dst_flume, started
async def open_fsp_chart(
self,