trustgraph/trustgraph-cli/trustgraph/cli/show_token_rate.py
cybermaggedon 9f2bfbce0c
Per-workspace queue routing for workspace-scoped services (#862)
Workspace identity is now determined by queue infrastructure instead of
message body fields, closing a privilege-escalation vector where a caller
could spoof workspace in the request payload.

- Add WorkspaceProcessor base class: discovers workspaces from config at
  startup, creates per-workspace consumers (queue:workspace), and manages
  consumer lifecycle on workspace create/delete events
- Roll out to librarian, flow-svc, knowledge cores, and config-svc
- Config service gets a dual-queue regime: a system queue for
  cross-workspace ops (getvalues-all-ws, bootstrapper writes to
  __workspaces__) and per-workspace queues for tenant-scoped ops, with
  workspace discovery from its own Cassandra store
- Remove workspace field from request schemas (FlowRequest,
  LibrarianRequest, KnowledgeRequest, CollectionManagementRequest) and
  from DocumentMetadata / ProcessingMetadata — table stores now accept
  workspace as an explicit parameter
- Strip workspace encode/decode from all message translators and gateway
  serializers
- Gateway enforces workspace existence: reject requests targeting
  non-existent workspaces instead of routing to queues with no consumer
- Config service provisions new workspaces from __template__ on creation
- Add workspace lifecycle hooks to AsyncProcessor so any processor can
  react to workspace create/delete without subclassing WorkspaceProcessor
2026-05-04 10:30:03 +01:00

119 lines
No EOL
2.7 KiB
Python

"""
Dump out a stream of token rates, input, output and total. This is averaged
across the time since tg-show-token-rate is started.
"""
import os
import requests
import argparse
import json
import time
default_metrics_url = "http://localhost:8088/api/metrics"
DEFAULT_TOKEN = os.getenv("TRUSTGRAPH_TOKEN", None)
class Collate:
def look(self, data):
return sum(
[
float(x["value"][1])
for x in data["data"]["result"]
]
)
def __init__(self, data):
self.last = self.look(data)
self.total = 0
self.time = 0
def record(self, data, time):
value = self.look(data)
delta = value - self.last
self.last = value
self.total += delta
self.time += time
return delta/time, self.total/self.time
def dump_status(metrics_url, number_samples, period, token=None):
input_url = f"{metrics_url}/query?query=input_tokens_total"
output_url = f"{metrics_url}/query?query=output_tokens_total"
headers = {}
if token:
headers["Authorization"] = f"Bearer {token}"
resp = requests.get(input_url, headers=headers)
obj = resp.json()
input = Collate(obj)
resp = requests.get(output_url, headers=headers)
obj = resp.json()
output = Collate(obj)
print(f"{'Input':>10s} {'Output':>10s} {'Total':>10s}")
print(f"{'-----':>10s} {'------':>10s} {'-----':>10s}")
for i in range(number_samples):
time.sleep(period)
resp = requests.get(input_url, headers=headers)
obj = resp.json()
inr, inl = input.record(obj, period)
resp = requests.get(output_url, headers=headers)
obj = resp.json()
outr, outl = output.record(obj, period)
print(f"{inl:10.1f} {outl:10.1f} {inl+outl:10.1f}")
def main():
parser = argparse.ArgumentParser(
prog='tg-show-token-rate',
description=__doc__,
)
parser.add_argument(
'-m', '--metrics-url',
default=default_metrics_url,
help=f'Metrics URL (default: {default_metrics_url})',
)
parser.add_argument(
'-p', '--period',
type=int,
default=1,
help=f'Metrics period (default: 1)',
)
parser.add_argument(
'-n', '--number-samples',
type=int,
default=100,
help=f'Metrics period (default: 100)',
)
parser.add_argument(
'-t', '--token',
default=DEFAULT_TOKEN,
help=f'Bearer token for authentication (default: TRUSTGRAPH_TOKEN env var)',
)
args = parser.parse_args()
try:
dump_status(**vars(args))
except Exception as e:
print("Exception:", e, flush=True)
if __name__ == "__main__":
main()