Fluksio/app
0
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Schedule a node across every machine, not just this one

Browse Source 
The engine answered "where does this node run" twice, in two ways that could
not see each other: a device sent it to a worker carrying that label, and
resources were counted against the engine's own cores. Declaring both meant the
second answer won and nothing was counted at all — which the data-science
getting-started page and the worked example both do.

One question now, in flow/placement.py: of every machine attached, which could
grant what this node asked for, and which of those has it free. The books move
onto each machine — one accountant per worker, built from the inventory it
reported — and the waiting moves above them, where one condition variable can
be woken by a release anywhere or by a worker attaching. Locks go one way:
placer, then a machine's books, never back.

So a node asking for a card now finds the box that has one, rather than being
clamped down to none and run here. When nothing can grant the ask at all it is
still cut down and run — a flow written on a cluster has to work on a laptop —
but the ceiling is one real machine now, since taking the largest of each
dimension separately can describe a machine nobody has.

Two things fixed on the way. A device on a connector node held every batch run
of its flow forever, waiting for a worker that could never run an entry point.
And `prefer` falling back to the engine skipped the books, so the fallback held
nothing.

The bench flow's node has taken a `params` argument that with_settings has not
forwarded for some time, so the benchmark could not run at all: 62 ms median
submit-to-result with this, against the 61 ms on record.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A6HeySA27EkGANZN95QySW
This commit is contained in:
stroblme
2026-08-27 08:49:36 +02:00
co-authored by Claude Opus 5
parent 1a9753fa9d
commit 6ff56533f5
14 changed files with 1214 additions and 326 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/fluksio/flow/controller.py
+80 -37
View File
@@ -62,6 +62,7 @@ from fluksio.flow.pipeline import (
ValidationIssue,
ValueSource,
)
from fluksio.flow.placement import Placer
from fluksio.flow.remote import RemoteWorkerHub
from fluksio.flow.resources import ResourceAccountant, derive_env
from fluksio.flow.schemas import (
@@ -413,6 +414,7 @@ class FlowController:
workers: PythonWorkerPool | None = None,
remote: RemoteWorkerHub | None = None,
resources: ResourceAccountant | None = None,
placer: Placer | None = None,
) -> None:
self.store = store
# Without a pool, python nodes are compiled and run in this process —
@@ -420,9 +422,11 @@ class FlowController:
self.workers = workers
# Workers on other hosts. A node without a device never touches it.
self.remote = remote
# What the machine has, for the nodes that say what they need. Without
# one, a declaration is recorded and nothing is held against it.
# This machine's own books. The pool sizes its fair share off them.
self.resources = resources
# Every machine there is, for the nodes that say what they need.
# Without one, a declaration is recorded and nothing is held against it.
self.placer = placer
self.state = state if state is not None else MemoryState()
self.events = events
self.max_workers = max_workers
@@ -865,9 +869,18 @@ class FlowController:
# Building
# -------------------------------------------------------------------------
def _allocated(
def _runs_elsewhere(self, node_def: NodeDef) -> bool:
"""Whether this node asks for more than this machine could ever give."""
if node_def.resources is None or self.placer is None:
return False
wanted = node_def.resources
return not self.placer.local.fits(wanted.cpus, wanted.gpus, wanted.ram or 0)
def _placed(
self,
wanted: Resources,
device: str | None,
policy: str,
owner: str,
local: str,
code: str,
@@ -877,32 +890,40 @@ class FlowController:
run_id: str,
on_event: Callable[[dict[str, Any]], None],
) -> Callable[..., Any]:
"""A call that holds its share of the machine while it runs.
"""A call that picks a machine and holds its share while it runs.
Which machine is decided per call rather than when the flow was built,
so a worker that attaches later is used without anything being rebuilt.
The order is load-bearing: the resources are claimed first, and only
then is a worker slot taken. The other way round, a node holding a slot
could sit waiting for cores that a node holding the cores cannot get a
slot to release.
The worker comes from the pool whose environment this allocation
derives, so what the node is told about its share is what the library
inside it reads at import — the only moment those variables are read.
What the allocation implies is handed to the process the node runs in,
here or on the worker, because a library reads those variables when it
is imported and never again.
"""
if self.workers is None or self.resources is None:
return self.workers.proxy( # type: ignore[union-attr]
owner,
local,
code,
node_id=node_id,
timeout=timeout,
run_id=run_id,
on_event=on_event,
)
accountant, pool = self.resources, self.workers
placer, pool = self.placer, self.workers
def call(**kwargs: Any) -> Any:
with accountant.claim(wanted, node=node_id, run=run_id) as allocation:
return pool.for_env(derive_env(wanted, allocation)).run(
with placer.claim(
wanted, device=device, policy=policy, node=node_id, run=run_id
) as (target, allocation):
env = derive_env(wanted, allocation)
if target.worker is None:
return pool.for_env(env).run(
owner,
local,
code,
kwargs,
node_id,
timeout,
run_id=run_id,
on_event=on_event,
)
return self.remote.run_on( # type: ignore[union-attr]
target.worker,
owner,
local,
code,
@@ -911,6 +932,7 @@ class FlowController:
timeout,
run_id=run_id,
on_event=on_event,
env=env,
)
return call
@@ -1003,6 +1025,11 @@ class FlowController:
problem = self.remote.compile(
node_def.device or "", owner, local, code
)
elif self._runs_elsewhere(node_def):
# Asks for more than this machine has, so it will run on
# one that has it. Same reason as a device: checking the
# import here would fail a node that is fine there.
problem = None
else:
problem = self.workers.compile(owner, local, code)
if problem:
@@ -1016,18 +1043,15 @@ class FlowController:
if node_def.timeout is not None
else settings.FLOW_NODE_TIMEOUT
)
function = self.workers.proxy(
owner,
local,
code,
node_id=node_id,
timeout=timeout,
run_id=run.run_id if run else "",
on_event=emissions.handle,
)
if node_def.resources is not None and not node_def.device:
function = self._allocated(
if node_def.resources is not None and self.placer is not None:
# Says how much of a machine it takes, so which machine
# and how much of it are one decision — including when
# it also names a device, which used to mean the two
# answers disagreed and nothing was accounted at all.
function = self._placed(
node_def.resources,
node_def.device,
node_def.device_policy,
owner,
local,
code,
@@ -1036,11 +1060,9 @@ class FlowController:
run_id=run.run_id if run else "",
on_event=emissions.handle,
)
if node_def.device and self.remote is not None:
# A node with a device runs on a worker carrying that
# label. Which worker is decided per call, so one that
# attaches after this flow was built is used without
# anything being rebuilt.
elif node_def.device and self.remote is not None:
# A device and nothing about size: the label alone
# decides, least busy first, as it always has.
function = self.remote.proxy(
node_def.device,
owner,
@@ -1051,9 +1073,30 @@ class FlowController:
run_id=run.run_id if run else "",
on_event=emissions.handle,
fallback=(
function if node_def.device_policy == "prefer" else None
self.workers.proxy(
owner,
local,
code,
node_id=node_id,
timeout=timeout,
run_id=run.run_id if run else "",
on_event=emissions.handle,
)
if node_def.device_policy == "prefer"
else None
),
)
else:
# Declares nothing: the shared pool, at no extra cost.
function = self.workers.proxy(
owner,
local,
code,
node_id=node_id,
timeout=timeout,
run_id=run.run_id if run else "",
on_event=emissions.handle,
)
# Outermost, so it sees the result whichever of the three
# above answered the call.
function = emissions.wrap(function)
backend/fluksio/flow/placement.py
+390
View File
@@ -0,0 +1,390 @@
"""Which machine a node runs on, and the waiting when none of them has room.
The engine used to answer this twice, in two ways that could not see each
other. A node with a ``device`` went to a worker carrying that label, chosen by
which of them had least in flight — a machine's *name*, never its size. A node
with ``resources`` was accounted against the engine's own cores, and only the
engine's: declaring both meant the second answer silently won and nothing was
counted at all.
One question, then, asked once here: of every machine attached — this one and
each worker — which could grant what this node asked for, and which of those
has it free right now. A worker reports its inventory when it attaches, so the
answer covers the whole installation rather than the host the engine happens to
be on.
Two properties are worth stating because they are what make the waiting safe:
*Waiting is central.* One condition variable, woken by a release on any machine
and by a worker attaching. A node waiting inside one machine's books could not
be woken by a second machine appearing, which is exactly the case a cluster is
for.
*Locks go one way.* This condition, then a machine's books, never the reverse —
:meth:`ResourceAccountant.try_take` does not block and
:meth:`ResourceAccountant.give_back` calls back out after dropping its own
lock. The other invariant the engine already had still holds too: the claim is
taken before a worker slot, so a node holding a slot can never be waiting on
resources that a node holding those resources cannot get a slot to release.
"""
from __future__ import annotations
import logging
import threading
import time
from collections.abc import Iterator
from contextlib import contextmanager
from dataclasses import dataclass
from typing import TYPE_CHECKING, Any
from fluksio.flow.events import EventBus
from fluksio.flow.resources import Allocation, ResourceAccountant
from fluksio.flow.schemas import Resources
if TYPE_CHECKING: # pragma: no cover - imported for types only
from fluksio.flow.provision import Provisioner
from fluksio.flow.remote import RemoteWorker, RemoteWorkerHub
logger = logging.getLogger(__name__)
#: How long a waiting node sleeps before looking again without being woken.
#: Everything that frees a machine wakes it, so this only catches a wakeup lost
#: to a race and gives a provisioned machine's timeout somewhere to be noticed.
WAKE_S = 5.0
@dataclass(frozen=True)
class Target:
"""One machine a node could run on: this engine, or an attached worker."""
name: str
accountant: ResourceAccountant
#: None is the engine itself, which runs nodes in its own worker pool.
worker: RemoteWorker | None = None
def carries(self, device: str) -> bool:
"""Whether a node asking for this device may run here.
The engine carries no label: a label names a machine somebody attached
for the purpose, and the engine is where the node would have run anyway.
"""
if self.worker is None:
return False
return device == self.name or device in self.worker.labels
@dataclass
class _Waiting:
node: str
since: float
reason: str
class Placer:
"""Every machine the engine can reach, and who is queued for one."""
def __init__(
self,
local: ResourceAccountant,
hub: RemoteWorkerHub | None = None,
events: EventBus | None = None,
) -> None:
self.local = local
self.hub = hub
self.events = events
self.provisioners: list[Provisioner] = []
self._condition = threading.Condition()
self._waiting: dict[int, _Waiting] = {}
self._ticket = 0
local.on_release = self.wake
# -- the machines ----------------------------------------------------------
def wake(self) -> None:
"""Something changed: a machine freed up, or one attached or left."""
for provisioner in self.provisioners:
provisioner.reconcile(self._attached_names())
with self._condition:
self._condition.notify_all()
def _attached_names(self) -> set[str]:
if self.hub is None:
return set()
return {worker.name for worker in self.hub.workers() if not worker.gone}
def targets(self) -> list[Target]:
"""This engine, and every worker attached right now."""
found = [Target(name="engine", accountant=self.local)]
if self.hub is not None:
for worker in self.hub.workers():
if worker.gone:
continue
# Assigned here rather than at attach so a worker that
# reconnects — a new object, with fresh books — is wired up
# without the hub having to know about placement at all.
worker.accountant.on_release = self.wake
found.append(
Target(
name=worker.name,
accountant=worker.accountant,
worker=worker,
)
)
return found
def _candidates(
self,
cpus: int,
gpus: int,
ram_mb: int,
device: str | None,
policy: str,
) -> list[Target]:
"""The machines that could run this node, best first."""
targets = self.targets()
remote = [target for target in targets if target.worker is not None]
engine = [target for target in targets if target.worker is None]
if device:
labelled = [target for target in remote if target.carries(device)]
fitting = self._fitting(labelled, cpus, gpus, ram_mb)
if fitting or policy != "prefer":
# `require` waits for the labelled machine even when the engine
# could take the node: the label is a statement about where
# this node is correct, not a preference.
return fitting
# `prefer` is what makes a flow work before the GPU box exists. It
# falls back only when nothing labelled could ever take the node —
# one that is merely busy is worth queueing for.
return self._fitting(engine, cpus, gpus, ram_mb)
# No label: the engine first when it fits, because a node that runs
# here costs no network and reuses a warm worker.
ordered = engine + sorted(
remote, key=lambda target: target.worker.in_flight if target.worker else 0
)
return self._fitting(ordered, cpus, gpus, ram_mb)
@staticmethod
def _fitting(
targets: list[Target], cpus: int, gpus: int, ram_mb: int
) -> list[Target]:
return [t for t in targets if t.accountant.fits(cpus, gpus, ram_mb)]
# -- what can actually be granted ------------------------------------------
def _shapes(self, device: str | None, policy: str) -> list[tuple[int, int, int]]:
"""Every machine's size, as something to measure a request against.
A provisioner's job shapes count: a machine it can start on demand is
one this installation has, even when nothing is attached yet.
"""
shapes = []
for target in self.targets():
if device and not target.carries(device):
# A `prefer` node may end up here, so the engine still counts.
if policy == "require" or target.worker is not None:
continue
books = target.accountant
shapes.append((books.cpus, books.gpus, books.ram_mb or 0))
for provisioner in self.provisioners:
shapes.extend(provisioner.shapes(device))
return shapes or [(self.local.cpus, self.local.gpus, self.local.ram_mb or 0)]
def _clamp(
self, wanted: Resources, node: str, device: str | None, policy: str
) -> tuple[int, int, int]:
"""What can be granted somewhere, which may be less than was asked.
A flow written against a sixty-four core box should still run on a
laptop — waiting forever for cores that do not exist is not a smaller
machine, it is a hung run. What it must not become is an ask no single
machine can grant: taking the largest of each dimension separately can
describe a machine that is not there, and waiting on *that* is the hung
run this exists to avoid. So the GPUs are settled first, and the rest is
measured only against machines that carry that many.
"""
ram = wanted.ram or 0
shapes = self._shapes(device, policy)
gpus = min(wanted.gpus, max(shape[1] for shape in shapes))
capable = [shape for shape in shapes if shape[1] >= gpus]
cpus = min(wanted.cpus, max(shape[0] for shape in capable))
# A machine that reported no memory is not a memory limit; only those
# that did have anything to say about it.
said = [shape[2] for shape in capable if shape[2]]
ram = min(ram, max(said)) if said and ram else ram
if (cpus, gpus, ram) != (wanted.cpus, wanted.gpus, wanted.ram or 0):
logger.warning(
"%s asked for %d cpu(s), %d gpu(s) and %s MB; "
"the largest machine here can give %d, %d and %s",
node or "a node",
wanted.cpus,
wanted.gpus,
wanted.ram or "no stated",
cpus,
gpus,
ram or "no stated",
)
return cpus, gpus, ram
# -- holding a machine for the length of a call ----------------------------
@contextmanager
def claim(
self,
wanted: Resources,
*,
device: str | None = None,
policy: str = "require",
node: str = "",
run: str = "",
) -> Iterator[tuple[Target, Allocation]]:
"""Hold a machine's share of itself for as long as this node runs."""
waited_from = time.monotonic()
ticket = 0
chosen: Target | None = None
allocation: Allocation | None = None
with self._condition:
while chosen is None:
cpus, gpus, ram = self._clamp(wanted, node, device, policy)
for target in self._candidates(cpus, gpus, ram, device, policy):
got = target.accountant.try_take(cpus, gpus, ram)
if got is not None:
chosen, allocation = target, got
break
if chosen is not None:
break
if not ticket:
ticket = self._announce(
node, run, waited_from, self._reason(cpus, gpus, device)
)
self._provision(cpus, gpus, ram, device)
self._condition.wait(timeout=WAKE_S)
if ticket:
self._waiting.pop(ticket, None)
if ticket:
logger.info(
"%s waited %.1fs and runs on '%s'",
node or "a node",
time.monotonic() - waited_from,
chosen.name,
)
assert allocation is not None
try:
yield chosen, allocation
finally:
# Every way a call can end comes through here: an answer, a
# timeout, a cancellation, or the worker going away mid-call.
chosen.accountant.give_back(allocation)
def _reason(self, cpus: int, gpus: int, device: str | None) -> str:
parts = []
if gpus:
parts.append(f"{gpus} gpu(s)")
parts.append(f"{cpus} cpu(s)")
wants = " and ".join(parts)
if device:
return f"waiting for a worker labelled '{device}' with {wants} free"
return f"waiting for {wants}"
def _provision(self, cpus: int, gpus: int, ram_mb: int, device: str | None) -> None:
"""Ask for a machine, if something here can start one."""
for provisioner in self.provisioners:
if provisioner.covers(cpus, gpus, ram_mb, device):
provisioner.provision(cpus, gpus, ram_mb, device)
return
def _announce(self, node: str, run: str, since: float, reason: str) -> int:
"""Say a node is queued, not stuck.
The failure this exists for looked identical to a hang: a run sitting
at `running` for twenty minutes with no error and no output. A node
waiting its turn has to say so somewhere a person will look.
"""
self._ticket += 1
ticket = self._ticket
self._waiting[ticket] = _Waiting(node=node, since=since, reason=reason)
logger.info("%s is %s", node or "a node", reason)
if self.events is not None:
self.events.publish(
{
"type": "node_queued",
"flow": node.split(".", 1)[0] if node else "",
"node": node,
"run": run,
"detail": reason,
"ts": time.time(),
}
)
return ticket
# -- what it looks like from outside ---------------------------------------
def satisfiable(self, needs: dict[str, Any] | None) -> str | None:
"""What a run would have to wait for, or None if it would not wait.
Only a genuine wait is worth holding a run for. A request nothing can
grant is cut down to what is here and runs anyway, so answering "no
machine has 64 cores" would hold a run that was about to work.
"""
if not needs:
return None
cpus = int(needs.get("cpus") or 1)
gpus = int(needs.get("gpus") or 0)
ram = int(needs.get("ram_mb") or 0)
device = needs.get("device") or None
for target in self.targets():
if device and not target.carries(device):
continue
if target.accountant.fits(cpus, gpus, ram):
# Something here could take it. Busy is queueing, not waiting.
return None
if not any(
provisioner.covers(cpus, gpus, ram, device)
for provisioner in self.provisioners
):
# Nothing attached fits and nothing can be started, so the node
# will be clamped onto what is here and run. Holding the run would
# be holding it for something that is not going to happen — and
# where the wait is a missing *label*, `required_labels` says so
# already.
return None
return f"{gpus} gpu(s) and {cpus} cpu(s)" if gpus else f"{cpus} cpu(s)"
def snapshot(self) -> dict[str, Any]:
"""Every machine, what is free of it, and who is queued.
A node waiting its turn looks exactly like a node that has hung — the
run sits at `running` and says nothing — so what is waiting, and for
what, has to be readable somewhere.
"""
targets = []
for target in self.targets():
entry: dict[str, Any] = {"target": target.name}
entry.update(target.accountant.snapshot())
if target.worker is not None:
entry["labels"] = sorted(target.worker.labels)
entry["in_flight"] = target.worker.in_flight
targets.append(entry)
with self._condition:
waiting = list(self._waiting.values())
local = self.local.snapshot()
return {
# The engine's own figures stay where they were, so a reader that
# only ever knew about one machine still finds them.
"cpus": local["cpus"],
"gpus": local["gpus"],
"waiting": [
{
"node": entry.node,
"reason": entry.reason,
"seconds": round(time.monotonic() - entry.since, 1),
}
for entry in waiting
],
"targets": targets,
"provisioners": [p.status() for p in self.provisioners],
}
backend/fluksio/flow/remote.py
+52 -1
View File
@@ -32,6 +32,7 @@ from dataclasses import dataclass
from typing import Any
from fluksio.flow.nodes.base import NodeOutputError
from fluksio.flow.resources import ResourceAccountant
from fluksio.flow.workers import NodeTimeout, RemoteError, _remote_class
logger = logging.getLogger(__name__)
@@ -106,6 +107,14 @@ class RemoteWorker:
self.labels = set(labels)
self.info = info or {}
self.inventory = inventory or WorkerInventory()
# This machine's books, held here so that a worker reconnecting — which
# is a new object — starts from a clean set rather than from counters
# the engine kept for a socket that is gone.
self.accountant = ResourceAccountant(
cpus=self.inventory.cpus,
gpus=self.inventory.gpus,
ram_mb=self.inventory.ram_mb,
)
self.attached_at = time.time()
self.last_seen = time.time()
self._send = send
@@ -251,9 +260,13 @@ class RemoteWorker:
class RemoteWorkerHub:
"""Every attached worker, and which of them a node may run on."""
def __init__(self) -> None:
def __init__(self, on_change: Callable[[], None] | None = None) -> None:
self._workers: dict[str, RemoteWorker] = {}
self._lock = threading.Lock()
#: Told when the set of attached workers changes, so whoever is waiting
#: for a machine can look again. Called outside the lock: it takes one
#: of its own, and the two are only ever taken in that order.
self.on_change = on_change
# -------------------------------------------------------------------------
# Attachment
@@ -270,6 +283,7 @@ class RemoteWorkerHub:
logger.info(
"Worker '%s' attached with labels %s", worker.name, sorted(worker.labels)
)
self._changed()
def detach(self, name: str) -> None:
with self._lock:
@@ -277,6 +291,11 @@ class RemoteWorkerHub:
if worker is not None:
worker.detach()
logger.info("Worker '%s' detached", name)
self._changed()
def _changed(self) -> None:
if self.on_change is not None:
self.on_change()
def workers(self) -> list[RemoteWorker]:
with self._lock:
@@ -326,6 +345,38 @@ class RemoteWorkerHub:
worker = self.pick(label)
if worker is None:
raise NoWorker(f"no worker labelled '{label}' is attached")
return self.run_on(
worker,
flow,
node,
source,
kwargs,
node_id,
timeout,
run_id=run_id,
on_event=on_event,
env=env,
)
def run_on(
self,
worker: RemoteWorker,
flow: str,
node: str,
source: str,
kwargs: dict[str, Any],
node_id: str,
timeout: float,
run_id: str = "",
on_event: Callable[[dict[str, Any]], None] | None = None,
env: dict[str, str] | None = None,
) -> Any:
"""Run this node on the worker the caller has already settled on.
The placer holds that machine's cores for the length of the call, so
resolving the label a second time here could send the work somewhere
else and leave the claim on a machine doing nothing.
"""
payload = {
"op": "run",
"call_id": f"{run_id}:{node_id}" if run_id else node_id,
backend/fluksio/flow/resources.py
+69 -124
View File
@@ -8,16 +8,18 @@ seconds. On a GPU the same shape is worse: three processes each preallocating
most of the card deadlock at zero utilisation, with nothing failing and nothing
to read.
:class:`ResourceAccountant` is the first half: a node that declares what it
needs waits until that much is free, the way it already waits for a worker
slot. :func:`derive_env` is the second: what it got is handed to the library as
the environment it reads at import, because that is the only moment those knobs
are read.
:class:`ResourceAccountant` is the first half: the books for one machine, which
hand out what is free and take it back. :func:`derive_env` is the second: what
a node got is handed to the library as the environment it reads at import,
because that is the only moment those knobs are read.
One accountant is one machine. Which machine a node goes to, and the waiting
when none of them has room, is :mod:`fluksio.flow.placement` — the books here
never block, so a caller holding several of them can ask each in turn.
Deliberately cooperative — nothing here is enforced with cgroups or rlimits, so
a node that ignores its share is only accounted for, not stopped. That is the
same trust the worker pool already extends to node code, and the enforcement
half belongs with the scheduler work this is the first step of.
same trust the worker pool already extends to node code.
"""
from __future__ import annotations
@@ -25,12 +27,9 @@ from __future__ import annotations
import logging
import os
import threading
import time
from collections.abc import Iterator
from contextlib import contextmanager
from collections.abc import Callable
from dataclasses import dataclass
from fluksio.flow.events import EventBus
from fluksio.flow.schemas import Resources
logger = logging.getLogger(__name__)
@@ -63,146 +62,92 @@ class Allocation:
cpus: int = 1
gpus: tuple[int, ...] = ()
@dataclass
class _Waiting:
node: str
since: float
reason: str
ram_mb: int = 0
class ResourceAccountant:
"""What is free on this machine, and who is waiting for it.
"""The books for one machine: what it holds, and what is free of it.
A plain condition variable over two counters. Blocking is the whole
mechanism: the caller is a node thread that would otherwise be executing,
and making it wait is the same backpressure that the worker pool's slot
queue already applies — one reason the two must always be taken in the same
order, resources first, so neither can be held while waiting for the other.
Nothing here blocks. A node that has to wait waits in the placer, which
holds one of these per machine and tries each in turn — a lock per machine
could not answer "is there room anywhere", and a node waiting inside one
machine's books could not be woken by another machine attaching.
``on_release`` is how it tells the placer something came free. It is called
outside this object's lock, deliberately: the placer takes its own lock
there, and the two are only ever taken in that one order.
"""
def __init__(
self, cpus: int = 0, gpus: int = 0, events: EventBus | None = None
self,
cpus: int = 0,
gpus: int = 0,
ram_mb: int | None = None,
on_release: Callable[[], None] | None = None,
) -> None:
self.cpus = cpus or machine_cpus()
self.gpus = max(0, gpus)
self.events = events
#: None is memory this machine did not report, which is allowed: it is
#: then not accounted rather than accounted as zero, since zero would
#: refuse every node that asks for any.
self.ram_mb = ram_mb
self.on_release = on_release
self._free_cpus = self.cpus
self._free_gpus = list(range(self.gpus))
self._waiting: dict[int, _Waiting] = {}
self._ticket = 0
self._condition = threading.Condition()
self._free_ram = ram_mb or 0
self._lock = threading.Lock()
# -- what a caller asks for ------------------------------------------------
def fits(self, cpus: int, gpus: int, ram_mb: int = 0) -> bool:
"""Whether this machine could ever grant that much, busy or not.
def _clamp(self, wanted: Resources, node: str) -> tuple[int, int]:
"""What can be granted here, which may be less than what was asked.
A flow written against a sixty-four core box should still run on a
laptop. Waiting forever for cores that do not exist is not a smaller
machine, it is a hung run.
The question a placer asks before it decides to wait: waiting for a
card that is merely busy is queueing, and waiting for one the machine
does not have is a hung run.
"""
cpus = min(wanted.cpus, self.cpus)
gpus = min(wanted.gpus, self.gpus)
if cpus != wanted.cpus or gpus != wanted.gpus:
logger.warning(
"%s asked for %d cpu(s) and %d gpu(s); this engine has %d and %d",
node or "a node",
wanted.cpus,
wanted.gpus,
self.cpus,
self.gpus,
)
return cpus, gpus
if cpus > self.cpus or gpus > self.gpus:
return False
return self.ram_mb is None or ram_mb <= self.ram_mb
@contextmanager
def claim(
self, wanted: Resources, node: str = "", run: str = ""
) -> Iterator[Allocation]:
"""Hold this node's share for as long as it runs."""
cpus, gpus = self._clamp(wanted, node)
allocation = self._take(cpus, gpus, node, run)
try:
yield allocation
finally:
self._give_back(allocation)
def _take(self, cpus: int, gpus: int, node: str, run: str) -> Allocation:
waited_from = time.monotonic()
ticket = 0
with self._condition:
while self._free_cpus < cpus or len(self._free_gpus) < gpus:
if not ticket:
reason = (
f"waiting for {cpus} cpu(s) ({self._free_cpus} free)"
if self._free_cpus < cpus
else f"waiting for {gpus} gpu(s) ({len(self._free_gpus)} free)"
)
ticket = self._announce(node, run, waited_from, reason)
self._condition.wait()
def try_take(self, cpus: int, gpus: int, ram_mb: int = 0) -> Allocation | None:
"""Take this much if it is free right now, or answer that it is not."""
with self._lock:
if self._free_cpus < cpus or len(self._free_gpus) < gpus:
return None
if self.ram_mb is not None and self._free_ram < ram_mb:
return None
self._free_cpus -= cpus
taken = tuple(self._free_gpus.pop(0) for _ in range(gpus))
if ticket:
self._waiting.pop(ticket, None)
if ticket:
logger.info(
"%s waited %.1fs for %d cpu(s) and %d gpu(s)",
node or "a node",
time.monotonic() - waited_from,
cpus,
gpus,
)
return Allocation(cpus=cpus, gpus=taken)
if self.ram_mb is not None:
self._free_ram -= ram_mb
return Allocation(cpus=cpus, gpus=taken, ram_mb=ram_mb)
def _give_back(self, allocation: Allocation) -> None:
with self._condition:
def give_back(self, allocation: Allocation) -> None:
"""Return what an execution held, however it ended."""
with self._lock:
self._free_cpus += allocation.cpus
self._free_gpus.extend(allocation.gpus)
self._free_gpus.sort()
self._condition.notify_all()
# -- what it looks like from outside ---------------------------------------
def _announce(self, node: str, run: str, since: float, reason: str) -> int:
"""Say a node is queued, not stuck.
The failure this exists for looked identical to a hang: a run sitting
at `running` for twenty minutes with no error and no output. A node
waiting its turn has to say so somewhere a person will look.
"""
self._ticket += 1
ticket = self._ticket
self._waiting[ticket] = _Waiting(node=node, since=since, reason=reason)
logger.info("%s is %s", node or "a node", reason)
if self.events is not None:
self.events.publish(
{
"type": "node_queued",
"flow": node.split(".", 1)[0] if node else "",
"node": node,
"run": run,
"detail": reason,
"ts": time.time(),
}
)
return ticket
if self.ram_mb is not None:
self._free_ram += allocation.ram_mb
if self.on_release is not None:
self.on_release()
def snapshot(self) -> dict[str, object]:
"""What is free and who is waiting, for the workers screen."""
with self._condition:
free_cpus, free_gpus = self._free_cpus, len(self._free_gpus)
"""What this machine holds and what is free of it."""
with self._lock:
free_cpus, free_gpus, free_ram = (
self._free_cpus,
len(self._free_gpus),
self._free_ram,
)
return {
"cpus": {"total": self.cpus, "free": free_cpus},
"gpus": {"total": self.gpus, "free": free_gpus},
"waiting": [
{
"node": entry.node,
"reason": entry.reason,
"seconds": round(time.monotonic() - entry.since, 1),
}
for entry in list(self._waiting.values())
],
"ram_mb": (
None
if self.ram_mb is None
else {"total": self.ram_mb, "free": free_ram}
),
}
backend/fluksio/flow/runs.py
+56 -8
View File
@@ -160,6 +160,17 @@ def _has_no_body(
return not store.has_node_source(flow, node.id, draft=draft)
def _runs_on_a_worker(node: NodeDef) -> bool:
"""Whether this node's body could go to a worker at all.
Only a node whose source travels can: a connector is an entry point loaded
in this process, so a device on one is a field nothing reads. Holding a run
for a worker that could never take that node is a run that never starts.
"""
node_type = NODE_TYPES.get(node.type)
return node_type is not None and node_type.has_source
def required_labels(flow: FlowDef) -> list[str]:
"""Worker labels this flow cannot run without.
@@ -171,11 +182,38 @@ def required_labels(flow: FlowDef) -> list[str]:
{
node.device
for node in flow.nodes
if node.device and node.device.strip() and node.device_policy == "require"
if node.device
and node.device.strip()
and node.device_policy == "require"
and _runs_on_a_worker(node)
}
)
def required_resources(flow: FlowDef) -> dict[str, Any] | None:
"""The largest single thing this flow needs a machine to have.
Dimension by dimension rather than per node, which is deliberately blunt:
it answers "could this run start at all", not "in what order". A node that
only *prefers* its device is left out — it runs here when nothing carries
the label, so it is never what a run is waiting for.
"""
cpus, gpus, ram = 0, 0, 0
device = ""
for node in flow.nodes:
wanted = node.resources
if wanted is None or not _runs_on_a_worker(node):
continue
cpus = max(cpus, wanted.cpus)
gpus = max(gpus, wanted.gpus)
ram = max(ram, wanted.ram or 0)
if node.device and node.device_policy == "require":
device = node.device
if not (cpus or gpus or ram):
return None
return {"cpus": cpus, "gpus": gpus, "ram_mb": ram, "device": device}
#: What a run's output is called from outside it: ``@run:<id>.<output>``.
RUN_REF_PREFIX = "@run:"
@@ -759,6 +797,7 @@ class RunService:
no_cache=no_cache,
status="queued",
labels=required_labels(flow),
needs=required_resources(flow),
created_at=datetime.now(UTC),
actor=actor,
idempotency_key=idempotency_key,
@@ -854,7 +893,7 @@ class RunService:
if not item.run_id:
self.queue.ack(item)
continue
missing = self._missing_labels(item.run_id)
missing = self._missing_requirements(item.run_id)
if missing:
# Left in the queue rather than failed: submitting a run
# before turning the GPU box on is a normal way to work, and
@@ -870,21 +909,30 @@ class RunService:
except RuntimeError:
logger.warning("Run %s not started: shutting down", item.run_id)
def _missing_labels(self, run_id: str) -> list[str]:
"""Worker labels this run needs that nothing attached carries."""
def _missing_requirements(self, run_id: str) -> list[str]:
"""What this run needs that nothing attached can give it yet."""
with Session(db_engine) as session:
run = session.get(Run, run_id)
needed = list(run.labels) if run else []
if not needed:
return []
needs = dict(run.needs) if run and run.needs else None
hub = self.controller.remote
available = hub.labels() | {w.name for w in hub.workers()} if hub else set()
# A node that only prefers its label runs locally instead, so it is not
# a reason to hold the run back; that is decided per node at call time.
return sorted(set(needed) - available)
missing = [f"labelled {label}" for label in sorted(set(needed) - available)]
placer = self.controller.placer
if placer is not None:
# Only a wait something is going to end — a machine that could take
# this is being started for it. An ask nothing can grant is cut down
# to what is here and runs, so holding the run would be holding it
# for something that is not coming.
short = placer.satisfiable(needs)
if short:
missing.append(f"with {short}")
return missing
def _waiting(self, run_id: str, missing: list[str]) -> None:
reason = f"Waiting for a worker labelled {', '.join(missing)}"
reason = f"Waiting for a worker {' and '.join(missing)}"
try:
with Session(db_engine) as session:
session.exec(
backend/fluksio/flow/schemas.py
+10
View File
@@ -60,6 +60,16 @@ class Resources(BaseModel):
"runs, which is what keeps two preallocating processes apart."
),
)
ram: int | None = Field(
default=None,
ge=1,
description=(
"Megabytes held for the whole execution. Counted against machines "
"that said how much they have, and ignored by those that did not — "
"which is a machine with nothing to say about memory, not one with "
"none."
),
)
env: dict[str, str] = Field(
default_factory=dict,
description=(