Arc vs QuestDB
Running QuestDB costs more to store the same data. On the ClickBench dataset, QuestDB holds 67.9 GiB while Arc holds it in 13.8 GiB of portable Parquet — 4.9x less at rest — and loads it 93x faster.
The Storage Bill: 4.9x More for the Same Data
Both databases were loaded with the identical ClickBench dataset — 99.9M rows, same columns. Arc compresses it to 13.8 GiB of Apache Parquet; QuestDB writes 67.9 GiB. On S3 or any object store, that is roughly five times the bytes you pay for, every month, for the same data.
| System | On-disk size | Relative | Format |
|---|---|---|---|
| Arc | 13.8 GiB | ×1.0 | Native Parquet, queryable in place |
| QuestDB | 67.9 GiB | ×4.93 | Native column format; Parquet via export |
Arc's data is standard Parquet you can query with DuckDB, Spark, or Snowflake directly — no export step. QuestDB stores in its own column format; reaching Parquet means an explicit export.
ClickBench Results
99.9M rows, 43 analytical queries. Arc runs true cold runs: service restart and OS cache flush before every query. Verify on benchmark.clickhouse.com →
Combined Score (lower is better)
| System | Machine | Score |
|---|---|---|
| Arc | c8g.metal-48xl | ×1.38 |
| Arc | c7a.metal-48xl | ×1.51 |
| Arc | c6a.4xlarge | ×2.38 |
| QuestDB | c7a.metal-48xl | ×2.64 |
| QuestDB | c6a.4xlarge | ×5.12 |
Cold Run (lower is better)
| System | Machine | Score |
|---|---|---|
| Arc | c8g.metal-48xl | ×1.15 |
| Arc | c6a.4xlarge | ×1.25 |
| Arc | c7a.metal-48xl | ×1.26 |
| QuestDB | c6a.4xlarge | ×5.33 |
| QuestDB | c7a.metal-48xl | ×7.45 |
Load Time (lower is better)
| System | Machine | Time to load 99.9M rows |
|---|---|---|
| Arc | c6a.4xlarge | 58s |
| Arc | c7a.metal-48xl | 68s |
| QuestDB | c7a.metal-48xl | 325s |
| QuestDB | c6a.4xlarge | 5,434s |
On the standard c6a.4xlarge machine, Arc loads the full dataset in 58 seconds; QuestDB takes over 90 minutes — 93x slower.
Two Different Bets
QuestDB is built for high-frequency ingestion of simple time-series. Arc is built for analytical depth and low storage cost on data you own. Where they overlap, the cost profile diverges.
Storage Cost
4.9x smaller footprint you own
Arc stores every table natively as compressed Apache Parquet on storage you control (local disk, S3, MinIO). The same dataset that fills 67.9 GiB in QuestDB fits in 13.8 GiB with Arc. On object storage that is a recurring bill you pay every month — and Arc's files are queryable in place by any Parquet tool, so there is no export step if you leave. QuestDB can produce Parquet too, but only through an explicit export from its own storage format.
Query Engine
Full DuckDB SQL vs. time-series SQL
Arc embeds DuckDB, a vectorized OLAP engine with the full analytical SQL surface — window functions, CTEs, and complex joins run without a performance penalty. QuestDB supports these too, plus its own time-series extensions like SAMPLE BY, but is tuned for flat time-series queries rather than deep analytical work.
Ingestion
Line Protocol on both sides
Both Arc and QuestDB accept InfluxDB Line Protocol, so existing Telegraf pipelines can point at either with a URL change. Arc also speaks a MessagePack columnar protocol sustained at 19.9M records/s, plus bulk CSV/Parquet import.
Deployment
Single Go binary, managed option available
Arc ships as a single Go binary and runs at the edge, on-prem, or in your own cloud, with Arc Enterprise Managed for teams that want it hosted for them. QuestDB's self-serve cloud was wound down; its managed path is now Enterprise BYOC, which requires an enterprise engagement rather than a self-serve signup.
Feature Comparison
| Feature | Arc | QuestDB |
|---|---|---|
| Standard SQL (window functions, CTEs, joins) | ✓ | ✓ |
| Native Parquet storage, no export step | ✓ | ✗ |
| Open source | ✓ | ✓ |
| InfluxDB Line Protocol ingestion | ✓ | ✓ |
| Compact storage (4.9x smaller on ClickBench) | ✓ | ✗ |
| Edge / single-binary deployment | ✓ | ✓ |
| Managed hosting option | ✓ | ✓ |
Both are open source; QuestDB is Apache 2.0, Arc is AGPL-3.0 with a commercial license available.
Frequently Asked Questions
Does QuestDB use more storage than Arc?
Yes. On the ClickBench dataset (99.9M rows), QuestDB stores 67.9 GiB while Arc stores the same data in 13.8 GiB of Apache Parquet — 4.9x less. On object storage that difference is a direct, recurring cost: you pay for roughly five times the bytes at rest with QuestDB.
How does Arc compare to QuestDB on ClickBench?
On the standard c6a.4xlarge machine, Arc leads QuestDB on combined score (×2.38 vs ×5.12), cold runs (×1.25 vs ×5.33), and load time (58s vs 5,434s). QuestDB is competitive on some individual hot-run queries. Full results are publicly verifiable on benchmark.clickhouse.com.
Can I migrate from QuestDB to Arc?
Yes. Both accept InfluxDB Line Protocol, so existing Telegraf and ILP writers can point at Arc with a URL change. QuestDB SQL maps to Arc's standard DuckDB SQL. The migration path is documented at basekick.net/migrate/questdb.
Does QuestDB support window functions and CTEs?
Yes, QuestDB supports window functions, CTEs, and joins. The practical difference is analytical depth and cost: QuestDB is tuned for high-frequency ingestion and flat time-series queries, while Arc runs the full DuckDB analytical SQL surface and stores data far more compactly.
Pricing
Start free with open source. Scale with enterprise features when you need them.
Open Source
- 19.9M records/sec ingestion
- Full SQL analytical engine
- Continuous queries + auto-compaction
- Open file format on S3, Azure, GCS, MinIO, local
- Docker and Kubernetes ready
- Community support (Discord)
Arc Enterprise Managed
Managed hosting, sized to your workload.
- Everything in Arc Enterprise
- Managed and operated by Basekick Labs
- Dedicated physical servers, sized to spec
- Daily backups to S3, monitoring, upgrades
- Migration support included
Enterprise
Starting price for up to 8 cores. Clustering, RBAC, and dedicated support.
- Everything in Open Source
- Horizontal clustering and HA
- Role-based access control (RBAC)
- Tiered storage automation
- Audit logging and query governance
- Dedicated support and SLAs
Enterprise Features
Clustering
Horizontal scaling with automatic data distribution. Query routing and load balancing across nodes.
Security
Fine-grained RBAC with database and table-level permissions. LDAP/SAML integration available.
Data Management
Automated retention policies, continuous queries for aggregation, and tiered storage for cost optimization.
Ready to handle billion-record workloads?
Deploy Arc in minutes. Own your data in open files on your storage. Use for analytics, observability, AI, IoT, or data warehousing.