InfluxDB 3 Core Benchmarks: QuestDB Comparison

This article presents benchmark results for InfluxDB 3 Core compared to QuestDB 9.2.2. InfluxDB 3 Core, which reached general availability in April 2025, is a complete rewrite of InfluxDB using Apache Arrow and DataFusion.

Key results: QuestDB ingests data 12x to 36x faster than InfluxDB 3 Core (up to 11.4M rows/sec vs 320K rows/sec), and runs analytical queries 17x to 418x faster depending on query type. InfluxDB 3 Core shows consistent ~320K rows/sec ingestion regardless of cardinality.

For InfluxDB v1/v2 benchmarks and detailed architecture comparison, see QuestDB vs InfluxDB.

About InfluxDB 3 Core

InfluxDB 3 Core represents a major architectural shift for InfluxDB:

• Complete rewrite in Rust (previously Go)
• Apache Arrow for in-memory data representation
• DataFusion query engine (Apache Arrow's SQL query engine)
• Parquet for storage format
• Arrow Flight for query protocol (gRPC-based)

This is a fundamentally different database from InfluxDB v1 and v2. The new architecture aims to improve analytical query performance through columnar processing.

Performance benchmarks

We use the open-source, industry-standard Time Series Benchmark Suite (TSBS) for all benchmarks. We extended the InfluxDB query runner to support v3 via Arrow Flight using the official influxdb3-go client library.

Hardware: AWS EC2 r8a.8xlarge (32 vCPU, 256 GB RAM, AMD EPYC), GP3 EBS storage (20,000 IOPS, 1 GB/s throughput)

Software:

• Ubuntu 22.04
• InfluxDB 3 Core (quay.io/influxdb/influxdb3-core:latest as of December 2025)
• QuestDB 9.2.2

Ingestion benchmark

We test a cpu-only scenario with two days of CPU data for various numbers of simulated hosts (100 to 1M). This tests how each database handles increasing data volumes and cardinality.

Example commands:

$ ./tsbs_generate_data --use-case="cpu-only" --seed=123 --scale=4000 \
    --timestamp-start="2016-01-01T00:00:00Z" \
    --timestamp-end="2016-01-03T00:00:00Z" \
    --log-interval="10s" --format="influx" > /tmp/influx_data

$ ./tsbs_load_influx --db-name=benchmark --file=/tmp/influx_data \
    --urls=http://localhost:8181 --workers=32

The results for ingestion with 32 workers:

Key observations:

• InfluxDB 3 Core shows constant ~320K rows/sec regardless of cardinality (100 to 1M hosts)
• QuestDB scales from 4M to 11.4M rows/sec, peaking at 100K hosts
• QuestDB is 12x to 36x faster than InfluxDB 3 Core on ingestion

Query performance

We tested standard TSBS query types against both databases. All queries run with a single worker to measure per-query latency.

Dataset: cpu-only scenario, 4,000 hosts, 10-second intervals, 2 days of data (~69M rows)

As part of the standard TSBS benchmark, we test several types of popular time series queries:

• single-groupby: Aggregate CPU metrics for random hosts over specified time ranges
• double-groupby: Aggregate across ALL hosts, grouped by host and time intervals
• high-cpu-all: Full table scan finding hosts with CPU utilization above threshold

To run the benchmark:

$ ./tsbs_generate_queries --use-case="devops" --seed=123 --scale=4000 \
    --timestamp-start="2016-01-01T00:00:00Z" \
    --timestamp-end="2016-01-03T00:00:00Z" \
    --queries=1000 --query-type="single-groupby-1-1-1" \
    --format="influx" > /tmp/influx_query

$ ./tsbs_run_queries_influx --file=/tmp/influx_query \
    --db-name=benchmark --workers=1

Single-groupby queries

Query naming: single-groupby-{metrics}-{hosts}-{hours} (e.g., 5-1-12 = 5 metrics, 1 host, 12-hour range)

Double-groupby queries

Heavy queries

Explaining performance differences

InfluxDB 3 Core observations

Ingestion: The consistent ~320K rows/sec throughput across all cardinality levels suggests the bottleneck is not related to series management (unlike v1/v2) but rather in the write path itself—possibly WAL handling, Parquet encoding, or HTTP request processing.

Queries: Results show consistent ~55-65 ms latency for simple 1-hour queries regardless of the number of metrics or hosts queried. This suggests a fixed per-query overhead, possibly from:

• Arrow Flight connection/protocol overhead
• DataFusion query planning
• Optimization opportunities in the new architecture

Longer time ranges (12-hour queries) scale linearly with data volume, reaching ~700-740 ms. The heavy analytical query (high-cpu-all) takes ~17 seconds.

Comparison with InfluxDB v1/v2

Interestingly, InfluxDB 3 Core shows slower performance than InfluxDB v1 and v2 on most workloads:

The new Arrow/DataFusion architecture is still maturing. Performance may improve in future releases as the codebase is optimized.

QuestDB performance

QuestDB maintains sub-2 ms latency on single-groupby queries and 40-58 ms on full-table double-groupby aggregations. Ingestion peaks at 11.4M rows/sec at 100K hosts. The columnar storage with SIMD-optimized aggregation functions provides consistent performance across query types.

Conclusion

InfluxDB 3 Core represents InfluxData's bet on the Apache Arrow ecosystem. While current performance lags behind both QuestDB and older InfluxDB versions, the architectural foundation (Arrow, DataFusion, Parquet) provides a path to improvement. Notably, the new architecture eliminates the cardinality-based performance degradation that plagued v1/v2.

QuestDB's purpose-built columnar engine with SIMD optimization delivers 12x to 36x faster ingestion and 17x to 418x faster query performance on these benchmarks. For production workloads requiring high-throughput ingestion and predictable low-latency queries, QuestDB remains the faster choice.

We'll continue to monitor InfluxDB 3 Core performance as new versions are released.

Performance summary

Ready to try QuestDB? Get started with the quickstart guide or join our Slack community.