InfluxDB

BenchBox supports InfluxDB 3.x for time series benchmarking workloads via FlightSQL.

Overview

InfluxDB 3.x is a time series database built on the FDAP stack (Apache Arrow, DataFusion, and Parquet). It provides native SQL support through the FlightSQL protocol, making it well-suited for benchmarking time series query performance.

Key Features

Native SQL Support: Query via FlightSQL protocol (Flux deprecated)
Time Series Optimized: Built for high-cardinality time series data
Arrow Data Format: Native Apache Arrow for efficient data transfer
Two Deployment Modes: Core (OSS) and Cloud (managed service)

Deployment Modes

Mode	Description	Use Case
Core	Self-hosted open source	Development, testing, local benchmarking
Cloud	Managed InfluxDB Cloud service	Production workloads, serverless benchmarking

Installation

Prerequisites

Python 3.11+
InfluxDB 3.x server (Core or Cloud)
Authentication token with read/write permissions

Install BenchBox with InfluxDB Support

uv add benchbox --extra influxdb

pip install benchbox[influxdb]

The first command installs the influxdb extra with uv. The second does the same with pip.

This installs:

influxdb3-python - Official InfluxDB 3.x Python client
pyarrow - Apache Arrow data handling

Verify Installation

from benchbox.platforms.influxdb import InfluxDBAdapter, INFLUXDB_AVAILABLE
print(f"InfluxDB support available: {INFLUXDB_AVAILABLE}")

Usage

CLI Usage

benchbox run --platform influxdb:cloud \
  --benchmark tsbs-devops \
  --platform-option host=us-east-1-1.aws.cloud2.influxdata.com \
  --platform-option token=$INFLUXDB_TOKEN \
  --platform-option org=my-org \
  --platform-option database=benchmarks

benchbox run --platform influxdb:core \
  --benchmark tsbs-devops \
  --platform-option host=localhost \
  --platform-option port=8086 \
  --platform-option token=$INFLUXDB_TOKEN \
  --platform-option database=benchmarks

The first command targets InfluxDB Cloud. The second targets a local InfluxDB Core server. All options are passed as --platform-option KEY=VALUE.

CLI Arguments

Option key	Default	Description
`host`	`localhost`	InfluxDB server hostname
`port`	`8086`	Server port
`token`	-	Authentication token (or set `INFLUXDB_TOKEN` env var)
`org`	-	Organization name
`database`	`benchbox`	Database (bucket) name
`deployment_mode`	`cloud`	Deprecated alias of `--platform influxdb:core\|cloud`; `mode` is an alias of this option
`ssl`	`true`	Use SSL/TLS connection. The CLI passes option values as strings, so disabling SSL from the CLI is not supported; use the Python API with `ssl=False` for plaintext local servers (see below)

Python API

from benchbox.platforms.influxdb import InfluxDBAdapter

adapter = InfluxDBAdapter(
    host="us-east-1-1.aws.cloud2.influxdata.com",
    token="your-token",
    org="your-org",
    database="benchmarks",
    deployment_mode="cloud",
)

adapter = InfluxDBAdapter(
    host="localhost",
    port=8086,
    token="your-token",
    database="benchmarks",
    deployment_mode="core",
    ssl=False,
)

connection = adapter.create_connection()

result = connection.execute("SELECT * FROM cpu LIMIT 10")
print(result)

connection.close()

The first adapter targets InfluxDB Cloud. The second targets InfluxDB Core running locally in Docker. The remaining lines create a connection, execute a query, and close the connection.

Supported Benchmarks

TSBS DevOps

The Time Series Benchmark Suite (TSBS) DevOps workload is the primary benchmark for InfluxDB:

benchbox run --platform influxdb --benchmark tsbs-devops --scale 1

TSBS DevOps simulates a DevOps monitoring scenario with:

CPU metrics: Usage, user, system, idle, etc.
Memory metrics: Total, available, used, etc.
Disk metrics: IOPS, throughput, latency
Network metrics: Bytes in/out, packets, errors

Query Types

Query Type	Description
Single-host	Query metrics for one host over time range
Groupby	Aggregate metrics grouped by time buckets
Lastpoint	Most recent metric value per host
High-CPU	Find hosts with CPU above threshold
Double-groupby	Group by multiple dimensions

Configuration

Environment Variables

export INFLUXDB_TOKEN="your-token-here"

Setting the authentication token in the environment is recommended for security.

Connection Configuration

config = {
    "host": "localhost",
    "port": 8086,
    "token": "your-token",
    "org": "your-org",
    "database": "benchmarks",
    "ssl": False,
    "deployment_mode": "core",
}

adapter = InfluxDBAdapter.from_config(config)

This is a full configuration example.

InfluxDB Core vs Cloud

InfluxDB Core (Open Source)

docker run -d \
  --name influxdb \
  -p 8086:8086 \
  -e DOCKER_INFLUXDB_INIT_MODE=setup \
  -e DOCKER_INFLUXDB_INIT_USERNAME=admin \
  -e DOCKER_INFLUXDB_INIT_PASSWORD=password123 \
  -e DOCKER_INFLUXDB_INIT_ORG=benchbox \
  -e DOCKER_INFLUXDB_INIT_BUCKET=benchmarks \
  -e DOCKER_INFLUXDB_INIT_ADMIN_TOKEN=my-token \
  influxdb:3.0

This starts InfluxDB Core with Docker.

Core limitations:

No data compaction for historical queries
No delete capabilities via SQL
Optimized for “leading edge” (recent) data

InfluxDB Cloud

Sign up at cloud2.influxdata.com
Create a bucket (database)
Generate an API token with read/write permissions
Use the provided host URL in your configuration

Performance Considerations

Leading Edge Queries: InfluxDB Core is optimized for recent data queries
Time Range Filters: Always include time bounds for best performance
Cardinality: InfluxDB handles high-cardinality tag values well
Aggregations: Time-bucketed aggregations are highly optimized

Troubleshooting

Connection Issues

connection = adapter.create_connection()
if connection.test_connection():
    print("Connection successful")
else:
    print("Connection failed")

This tests the connection.

Common Errors

Error	Cause	Solution
`ConnectionError: Failed to connect`	Wrong host/port	Verify server is running and accessible
`No InfluxDB client library`	Missing dependency	Run `uv add influxdb3-python`
`Authentication failed`	Invalid token	Check token has correct permissions

Debug Logging

import logging
logging.getLogger("benchbox.platforms.influxdb").setLevel(logging.DEBUG)

Search docs and blog