fluxmeter/Makefile at main · 10kshuaizhang/fluxmeter · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
.PHONY: build demo demo-proof demo-gateway demo-record start start-saas start-benchmark stop stop-saas clean generate benchmark correctness-bench validate-spec load-test load-test-quick http-load-test http-load-test-single http-load-test-batch test-e2e test-unit test-java test-cold-store apply-cold-store-init

JAR = $(shell ls -t build/libs/fluxmeter-*.jar 2>/dev/null | head -1)

# Build the Flink metering engine embedded in the runtime image.
build:
	./gradlew shadowJar

# One-command demo for the only architecture.
demo: start
	@echo ""
	@echo "==================================="
	@echo " FluxMeter 4.7 — HTTP → Kafka → Flink → Redis"
	@echo "==================================="
	@echo " API:           http://localhost:8000/docs"
	@echo " Intelligence:  http://localhost:8000/docs#/intelligence"
	@echo " Gateway:       http://localhost:8080/v1/chat/completions"
	@echo " Grafana:       http://localhost:3000 (admin/fluxmeter)"
	@echo ""
	@echo " Flink UI:      http://localhost:8081"
	@echo " Record demo.gif: make demo-record  (requires vhs)"
	@echo ""
	@echo " Gateway example:"
	@echo "   curl localhost:8080/v1/chat/completions \\"
	@echo "     -H 'X-FluxMeter-Customer-Id: cust_1' -H 'Authorization: Bearer \$$OPENAI_API_KEY' \\"
	@echo "     -d '{\"model\":\"gpt-4o-mini\",\"messages\":[{\"role\":\"user\",\"content\":\"hi\"}]}'"
	@echo "==================================="

# Re-record demo.gif from demo.tape (brew install vhs)
demo-record:
	vhs demo.tape

# Gateway mock self-check (no live OpenAI)
demo-gateway:
	PYTHONPATH=api python demos/gateway_demo.py

# Deterministic reserve → meter → kill → audit proof (no provider key).
demo-proof: build
	docker compose -f docker-compose.yml -f docker-compose.benchmark.yml -f docker-compose.proof.yml up -d --build
	./scripts/apply-cold-store-init.sh
	@set -e; trap 'docker compose -f docker-compose.yml -f docker-compose.benchmark.yml up -d --no-deps --force-recreate gateway >/dev/null' EXIT; python -u demos/reserve_meter_kill_audit_demo.py

# Build and start Kafka, Flink, Redis, API, Gateway, webhook worker, and Grafana.
start: build
	docker compose up -d --build
	@echo "FluxMeter started. The Flink job is submitted automatically."

start-benchmark: build
	docker compose -f docker-compose.yml -f docker-compose.benchmark.yml up -d --build

# Stop everything
stop:
	docker compose down 2>/dev/null || true
	docker compose -f docker-compose.yml -f docker-compose.saas.yml down 2>/dev/null || true
	docker compose -f docker-compose.yml -f docker-compose.benchmark.yml down 2>/dev/null || true

# Clean build artifacts and containers
clean: stop
	./gradlew clean
	docker compose down -v 2>/dev/null || true
	docker compose -f docker-compose.yml -f docker-compose.saas.yml down -v 2>/dev/null || true
	docker compose -f docker-compose.yml -f docker-compose.benchmark.yml down -v 2>/dev/null || true

# Validate open spec artifacts
validate-spec:
	./scripts/validate-spec.sh

test-e2e:
	pip install -q -r tests/requirements.txt
	pytest tests/test_integration.py -v --timeout=300
	pytest tests/test_e2e_v2.py -v --timeout=300 -m v2

test-unit:
	pip install -q -r tests/requirements.txt
	pytest tests/test_auth_unit.py tests/test_billing_export.py tests/test_billing_export_partners.py \
		tests/test_hierarchy_reserve.py tests/test_api_key_budget.py tests/test_billing_dims.py \
		tests/test_control_plane_models.py tests/test_tenant_keys.py \
		tests/test_pricing_loader.py tests/test_pricing_validate.py \
		tests/test_rerate_tier.py tests/test_phase2_billing.py tests/test_gateway.py \
		tests/test_ingestion_contract.py tests/test_reservation_expiry.py \
		tests/test_webhook_worker.py tests/test_deep_modules.py tests/test_proof_demo.py \
		tests/test_custody_unit.py -v --timeout=60
	PYTHONPATH=sdk/python pytest sdk/python/tests -q
	cd sdk/js && npm run build
	./gradlew test -q

test-java:
	./gradlew test

# Staged internal engine load test (10K → 1M eps bursts)
load-test:
	./scripts/load-test.sh

# Quick load test (10K–500K only)
load-test-quick:
	QUICK=1 ./scripts/load-test.sh

# Formal HTTP gates: 5m warm-up + 30m measured, typical payload, open-loop.
# Run both targets even if one fails: make -k http-load-test
HTTP_LOAD_SINGLE_TARGET_EPS ?= 10050
HTTP_LOAD_SINGLE_CONCURRENCY ?= 4000
HTTP_LOAD_SINGLE_MAX_P50_MS ?= 50
HTTP_LOAD_SINGLE_MAX_P99_MS ?= 200

http-load-test: http-load-test-single http-load-test-batch

http-load-test-single: build
	java -cp $(JAR) io.fluxmeter.generator.HttpLoadGenerator --url http://127.0.0.1:8000/ingest --target-eps $(HTTP_LOAD_SINGLE_TARGET_EPS) --warmup-seconds 300 --duration-seconds 1800 --concurrency $(HTTP_LOAD_SINGLE_CONCURRENCY) --min-eps 10000 --max-p50-ms $(HTTP_LOAD_SINGLE_MAX_P50_MS) --max-p99-ms $(HTTP_LOAD_SINGLE_MAX_P99_MS)

http-load-test-batch:
	python3 scripts/http-load-test.py --mode batch --batch-size 1000 --target-eps 100000 --processes 1 --concurrency 100 --min-eps 100000 --max-p99-ms 500

# ADR-025 ClickHouse cold store DDL (benchmark overlay must be up)
apply-cold-store-init:
	chmod +x scripts/apply-cold-store-init.sh
	./scripts/apply-cold-store-init.sh

# ADR-025 acceptance A1–A7 (needs kafka + clickhouse from start-benchmark)
test-cold-store:
	chmod +x scripts/test-cold-store.sh scripts/apply-cold-store-init.sh
	./scripts/apply-cold-store-init.sh
	./scripts/test-cold-store.sh

# Run the baseline comparison (Flink vs ClickHouse)
benchmark:
	./baseline/benchmark.sh

# Known-event correctness + Flink checkpoint health
correctness-bench:
	chmod +x scripts/correctness-bench.sh
	./scripts/correctness-bench.sh

# Trusted operator load generator (benchmark overlay exposes Kafka locally).
generate:
	KAFKA_BROKERS=localhost:9094 \
	NUM_CUSTOMERS=10000 \
	NUM_THREADS=8 \
	TARGET_EPS=1000000 \
	java -cp $(JAR) io.fluxmeter.generator.LoadGenerator

start-saas: build
	docker compose -f docker-compose.yml -f docker-compose.saas.yml up -d --build
	@echo "SaaS stack started. API :8000, Control Plane :8001, Grafana :3000"

stop-saas:
	docker compose -f docker-compose.yml -f docker-compose.saas.yml down