set a default TTL for query plans #4588

Performance

Workflow

Name: github-event-v52k6
Creation timestamp: 2024-04-05T13:48:57Z
Archived workflow https://argo.starstuff.dev/archived-workflows/perf-testing?name=github-event-v52k6

All simulations passed

Simulations

	simulation	scenario	rps	minRps	maxRps	success	min_latency_ms	max_latency_ms	mean_latency_ms
0	const	basic	4009	3800	4200	True	5	16	9.016529
1	step	basic	7642	7000	9000	True	5	515	29.19883

Tests

const - basic

step - basic

const - basic

definition

attack:
  duration: PT2M
  type: constant
  users: 4000
checks:
  maxFailures: 10000
  maxRps: 4200
  minRps: 3800
operation:
  breadth: 2
  depth: 5
  fields: 100
  name: basic
  subgraphResponseFieldSize: 32
  timeout: PT0.5S
  type: query
protocol:
  http2: false
  shareConnections: true
warmup:
- duration: PT10S
  endUsers: 4000
  startUsers: 0
  type: ramp
- duration: PT10S
  type: constant
  users: 4000

Gatling

/tmp/ipykernel_150/2413066115.py:42: FutureWarning: 'S' is deprecated and will be removed in a future version, please use 's' instead.
  all = all.groupby([pd.Grouper(key='timestamp', freq='1S'), 'metric']).max().reset_index()

Node stats

step - basic

definition

attack:
  incrementUsers: 100
  rampDuration: PT1S
  startUsers: 6000
  stepDuration: PT10S
  times: 200
  type: step
checks:
  maxFailures: 10000
  maxRps: 9000
  minRps: 7000
operation:
  breadth: 2
  depth: 5
  fields: 100
  name: basic
  subgraphResponseFieldSize: 32
  timeout: PT0.5S
  type: query
protocol:
  http2: false
  shareConnections: true
warmup:
- duration: PT10S
  endUsers: 6000
  startUsers: 0
  type: ramp
- duration: PT10S
  type: constant
  users: 6000

Gatling

/tmp/ipykernel_150/2413066115.py:42: FutureWarning: 'S' is deprecated and will be removed in a future version, please use 's' instead.
  all = all.groupby([pd.Grouper(key='timestamp', freq='1S'), 'metric']).max().reset_index()