Skip to main content

6 posts tagged with "performance"

Performance optimization and analysis

View All Tags

Spice v1.9.0-rc.1 (Nov 4, 2025)

· 16 min read
William Croxson
William Croxson
Senior Software Engineer at Spice AI

This is the first release candidate for v1.9.0, which introduces Cayenne, a new high-performance data accelerator built on the Vortex columnar format that delivers DuckDB-comparable performance without scaling limitations. This release also upgrades to DataFusion v50 for improved query performance, expands search capabilities with full-text search on views and multi-column embeddings, includes significant DynamoDB and DuckDB accelerator improvements, and delivers security and reliability enhancements.

What's New in v1.9.0-rc.1​

Cayenne Data Accelerator (Alpha)​

Introducing Cayenne: SQL as an Acceleration Format: A new high-performance data accelerator that simplifies multi-file data acceleration by using an embedded database (SQLite) for metadata while storing data in the Vortex columnar format. Cayenne delivers query and ingestion performance comparable or better to DuckDB's file-based acceleration without DuckDB's memory overhead and the scaling challenges of single DuckDB files.

Cayenne uses SQLite to manage acceleration metadata (schemas, snapshots, statistics, file tracking) through simple SQL transactions, while storing actual data in Vortex's compressed columnar format. This architecture provides:

Key Features:

  • SQLite + Vortex Architecture: All metadata is stored in SQLite tables with standard SQL transactions, while data lives in Vortex's compressed, chunked columnar format designed for zero-copy access and efficient scanning.
  • Simplified Operations: No complex file hierarchies, no JSON/Avro metadata files, no separate catalog servers—just SQL tables and Vortex data files. The entire metadata schema is intentionally simple for maximum reliability.
  • Fast Metadata Access: Single SQL query retrieves all metadata needed for query planning—no multiple round trips to storage, no S3 throttling, no reconstruction of metadata state from scattered files.
  • Efficient Small Changes: Dramatically reduces small file proliferation. Snapshots are just rows in SQLite tables, not new files on disk. Supports millions of snapshots without performance degradation.
  • High Concurrency: Changes consist of two steps: stage Vortex files (if any), then run a single SQL transaction. Much faster conflict resolution and support for many more concurrent updates than file-based formats.
  • Advanced Data Lifecycle: Full ACID transactions, delete support, and retention SQL execution on refresh commit.

Example Spicepod.yml configuration:

datasets:
  - from: s3:my_table
    name: accelerated_data
    acceleration:
      enabled: true
      engine: cayenne
      retention:
        sql: DELETE FROM accelerated_data WHERE created_at < NOW() - INTERVAL '30 days'

Note, the Cayenne Data Accelerator is in Alpha with limitations.

For more details, refer to the Cayenne Documentation, the Vortex project, and the DuckLake announcement that partly inspired this design.

DataFusion v50 Upgrade​

Spice.ai is built on the DataFusion query engine. The v50 release brings significant performance improvements and enhanced reliability:

Performance Improvements 🚀:

  • Dynamic Filter Pushdown: Enhanced dynamic filter pushdown for custom ExecutionPlans, ensuring filters propagate correctly through all physical operators for improved query performance.
  • Partition Pruning: Expanded partition pruning support ensures that unnecessary partitions are skipped when filters are not used, reducing data scanning overhead and improving query execution times.

Bug Fixes & Reliability: Resolved issues with partition name validation and empty execution plans when vector index lists are empty. Fixed timestamp support for partition expressions, enabling better partitioning for time-series data.

See the Apache DataFusion 50.0.0 Release for more details.

DynamoDB Data Connector Improvements​

Improved Query Performance: The DynamoDB Data Connector now includes improved filter handling for edge cases, parallel scan support for faster data ingestion, and better error handling for misconfigured queries. These improvements enable more reliable and performant access to DynamoDB data.

Example Spicepod.yml configuration:

datasets:
  - from: dynamodb:my_table
    name: ddb_data
    params:
      scan_segments: 10 # Default `auto` which calculates optimal segments based on number of rows

Search & Embeddings Enhancements​

Full-Text Search on Views: Full-text search indexes are now supported on views, enabling advanced search scenarios over pre-aggregated or transformed data. This extends the power of Spice's search capabilities beyond base datasets.

Multi-Column Embeddings on Views: Views now support embedding columns, enabling vector search and semantic retrieval on view data. This is useful for search over aggregated or joined datasets.

Vector Engines on Views: Vector search engines are now available for views, enabling similarity search over complex queries and transformations.

Example Spicepod.yml configuration:

views:
  - name: aggregated_reviews
    sql: SELECT review_id, review_text FROM reviews WHERE rating > 4
    embeddings:
      - column: review_text
        model: openai:text-embedding-3-small

DuckDB Accelerator Improvements​

Parquet Buffering for Partitioned Writes: DuckDB partitioned writes in table mode now support Parquet buffering, reducing memory usage and improving write performance for large datasets.

Retention SQL on Refresh Commit: DuckDB accelerations now support running retention SQL on refresh commit, enabling automatic data cleanup and lifecycle management during refresh operations.

UTC Timezone for DuckDB: DuckDB now uses UTC as the default timezone, ensuring consistent behavior for time-based queries across different environments.

Example Spicepod.yml configuration:

datasets:
  - from: s3://my_bucket/large_table/
    name: partitioned_data
    acceleration:
      enabled: true
      engine: duckdb
      mode: file
      retention:
        sql: DELETE FROM partitioned_data WHERE event_time < NOW() - INTERVAL '7 days'

Query Performance Optimizations​

Optimized Prepared Statements: Prepared statement handling has been optimized for better performance with parameterized queries, reducing planning overhead and improving execution time for repeated queries.

Large RecordBatch Chunking: Large Arrow RecordBatch objects are now automatically chunked to control memory usage during query execution, preventing memory exhaustion for queries returning large result sets.

Security & Reliability Improvements​

Enhanced HTTP Client Security: HTTP client usage across the runtime has been hardened with improved TLS validation, certificate pinning for critical endpoints, and better error handling for network failures.

ODBC Connector Improvements: Removed unwrap calls from the ODBC connector, improving error handling and reliability. Fixed secret handling and Kubernetes secret integration.

CLI Permissions Hardening: Tightened file permissions for the CLI and install script, ensuring secure defaults for configuration files and credentials.

Oracle Instant Client Pinning: Oracle Instant Client downloads are now pinned to specific SHAs, ensuring reproducible builds and preventing supply chain attacks.

Observability & Tracing​

DataFusion Log Emission: The Spice runtime now emits DataFusion internal logs, providing deeper visibility into query planning and execution for debugging and performance analysis.

AI Completions Tracing: Fixed tracing so that ai_completions operations are correctly parented under sql_query traces, improving observability for AI-powered queries.

Git Data Connector (Alpha)​

Version-Controlled Data Access: The new Git Data Connector (Alpha) enables querying datasets stored in Git repositories. This connector is ideal for use cases involving configuration files, documentation, or any data tracked in version control.

Example Spicepod.yml configuration:

datasets:
  - from: git:https://github.com/myorg/myrepo
    name: git_metrics
    params:
      file_format: csv

For more details, refer to the Git Data Connector Documentation.

Additional Improvements & Bug Fixes​

  • Reliability: Fixed refresh worker panics with recovery handling to prevent runtime crashes during acceleration refreshes.
  • Reliability: Improved error messages for missing or invalid spicepod.yaml files, providing actionable feedback for misconfiguration.
  • Reliability: Fixed DuckDB metadata pointer loading issues for snapshots.
  • Performance: Ensured ListingTable partitions are pruned correctly when filters are not used.
  • Reliability: Fixed vector dimension determination for partitioned indexes.
  • Search: Fixed casing issues in Reciprocal Rank Fusion (RRF) for hybrid search queries.
  • Search: Fixed search field handling as metadata for chunked search indexes.
  • Validation: Added timestamp support for partition expressions.
  • Validation: Fixed regexp_match function for DuckDB datasets.
  • Validation: Fixed partition name validation for improved reliability.

Contributors​

Breaking Changes​

No breaking changes.

Cookbook Updates​

No major cookbook updates.

The Spice Cookbook includes 81 recipes to help you get started with Spice quickly and easily.

Upgrading​

To upgrade to v1.9.0-rc.1, use one of the following methods:

CLI:

spice upgrade

Homebrew:

brew upgrade spiceai/spiceai/spice

Docker:

Pull the spiceai/spiceai:1.9.0-rc.1 image:

docker pull spiceai/spiceai:1.9.0-rc.1

For available tags, see DockerHub.

Helm:

helm repo update
helm upgrade spiceai spiceai/spiceai

AWS Marketplace:

🎉 Spice is now available in the AWS Marketplace!

What's Changed​

Changelog​

Spice v1.8.3 (Oct 27, 2025)

· 5 min read
David Stancu
David Stancu
Principal Software Engineer at Spice AI

Announcing the release of Spice v1.8.3! ⚡

Spice v1.8.3 is a patch release focused on performance, reliability, and observability. This release delivers optimizations for DuckDB acceleration, parameterized queries, and query plans. A new opt-in dedicated thread pool for queries is now in preview.

What's New in v1.8.3​

DuckDB Data Accelerator Improvements​

  • Connection Pool Sizing: The DuckDB accelerator now supports a configurable connection_pool_size parameter, supporting fine-grained control over concurrent query execution. This enables tuning for high-concurrency workloads and improved resource utilization.

Example Spicepod.yaml snippet:

datasets:
  - from: postgres:my_table
    name: my_table
    acceleration:
      enabled: true
      engine: duckdb
      params:
        connection_pool_size: 10
  • Automatic Statistics Recomputation: The new on_refresh_recompute_statistics parameter, on by default, triggers automatic ANALYZE execution after refreshes. This keeps DuckDB optimizer statistics up-to-date, ensuring efficient query plans and optimal performance.

Example Spicepod.yaml snippet:

datasets:
  - from: postgres:my_table
    name: my_table
    acceleration:
      enabled: true
      engine: duckdb
      params:
        on_refresh_recompute_statistics: disabled # default enabled

Task History SQL Query Plan Capture & Configuration​

Spice now supports automated SQL query plan capture and store (via EXPLAIN or EXPLAIN ANALYZE) in the task history, enabling deeper analysis and debugging of query execution. This feature is configurable, supporting control of which queries are included based on duration thresholds and plan type.

  • New Configuration Options:
    • task_history.captured_plan: Controls which plan is captured (none, explain, or explain analyze). Default none.
    • task_history.min_sql_duration: Minimum query duration before a plan is captured.
    • task_history.min_plan_duration: Minimum plan execution duration before a plan is captured.

Example spicepod.yaml snippet:

runtime:
  task_history:
    captured_plan: explain analyze
    min_sql_duration: 5s
    min_plan_duration: 10s

Query plans are captured asynchronously to avoid blocking query execution. The result of the plan is stored in the standard sql_query output in the task history.

Learn more in the Task History Documentation.

Query Performance Optimizations​

  • Optimized Prepared Statements (Parameterized Queries): Prepared statement caching for parameterized SQL queries has been improved, reducing planning overhead for repeated queries with different parameters. This results in faster execution and lower latency for workloads that reuse query structures.

  • Limit Pushdown via BytesProcessedExec: Introduces the BytesProcessedExec physical operator, enabling limit pushdown for large datasets. This optimization reduces the amount of data processed and improves top-k query performance.

Dedicated Query Thread Pool (Opt-In)​

Spice now supports running query execution and accelerated refreshes on a dedicated thread pool, separate from the HTTP server. This prevents heavy query workloads from slowing down API responses, keeping health and readiness checks fast. Opt-In for v1.8.3: This feature is opt-in for this release and will become enabled by default (opt-out) in v1.9.

Example Spicepod.yaml snippet:

runtime:
  params:
    dedicated_thread_pool: sql_engine # Default: disabled

Validation & Reliability Improvements​

  • Selective Evaluation Scorer Loading: Evaluation scorers are now loaded only when evaluation is explicitly defined, reducing unnecessary initialization and improving startup performance.

  • Improved Error Reporting: Enhanced error messages for misconfigured full-text search (FTS) on datasets and views, providing actionable feedback for configuration issues.

REPL & Usability​

  • Execution Time Display: The Spice REPL now displays query execution time even when queries return no results, improving user feedback and diagnostics.

Contributors​

Breaking Changes​

No breaking changes.

Cookbook Updates​

No major cookbook updates.

The Spice Cookbook includes 81 recipes to help you get started with Spice quickly and easily.

Upgrading​

To upgrade to v1.8.3, use one of the following methods:

CLI:

spice upgrade

Homebrew:

brew upgrade spiceai/spiceai/spice

Docker:

Pull the spiceai/spiceai:1.8.3 image:

docker pull spiceai/spiceai:1.8.3

For available tags, see DockerHub.

Helm:

helm repo update
helm upgrade spiceai spiceai/spiceai

AWS Marketplace:

🎉 Spice is now available in the AWS Marketplace!

What's Changed​

Changelog​

Spice v1.0.3 (Feb 10, 2025)

· 3 min read
Phillip LeBlanc
Phillip LeBlanc
Co-Founder and CTO of Spice AI

Announcing the release of Spice v1.0.3 🛠️

Spice v1.0.3 provides several bug fixes, including a fix for the initial data load period when a retention policy has been set, and a new unsupported_type_action: string parameter to auto-convert unsupported types to strings.

Highlights in v1.0.3​

  • PostgreSQL Data Connector: New unsupported_type_action: string parameter that auto-converts unsupported types such as JSONB to strings.

Contributors​

  • @phillipleblanc
  • @Sevenannn
  • @sgrebnov
  • @peasee
  • @Jeadie
  • @lukekim

Breaking Changes​

No breaking changes.

Cookbook Updates​

Upgrading​

To upgrade to v1.0.3, use one of the following methods:

CLI:

spice upgrade

Homebrew:

brew upgrade spiceai/spiceai/spice

Docker:

Pull the spiceai/spiceai:1.0.3 image:

docker pull spiceai/spiceai:1.0.3

For available tags, see DockerHub.

Helm:

helm repo update
helm upgrade spiceai spiceai/spiceai

What's Changed​

Dependencies​

No major dependency changes.

Changelog​

- For local models, use 'content=""' instead of None by @Jeadie and @phillipleblanc in https://github.com/spiceai/spiceai/pull/4646
- Perplexity Sonar LLM component by @Jeadie and @lukekim in https://github.com/spiceai/spiceai/pull/4673
- Update async openai fork & support reasoning effort parameter by @Sevenannn and @phillipleblanc in https://github.com/spiceai/spiceai/pull/4679
- Web search tool by @Jeadie and @lukekim in https://github.com/spiceai/spiceai/pull/4687
- Setup tpc-extension by @ewgenius and @phillipleblanc in https://github.com/spiceai/spiceai/pull/4690
- fix: Use PostgreSQL interval style for Spice.ai by @peasee and @phillipleblanc in https://github.com/spiceai/spiceai/pull/4716
- Fix spice upgrade command by @Sevenannn and @sgrebnov in https://github.com/spiceai/spiceai/pull/4699
- Fix bug: Ensure refresh only retrieves data within the retention period by @sgrebnov and @phillipleblanc in https://github.com/spiceai/spiceai/pull/4717
- Implement unsupported_type_action: string for Postgres JSONB support by @phillipleblanc in https://github.com/spiceai/spiceai/pull/4719
- Fix the get latest release logic by @Sevenannn and @phillipleblanc in https://github.com/spiceai/spiceai/pull/4721
- add 'accelerated_refresh' to 'spice trace' allowlist by @Jeadie and @phillipleblanc in https://github.com/spiceai/spiceai/pull/4711
- Update version to 1.0.3 by @phillipleblanc in https://github.com/spiceai/spiceai/pull/4731
- Truncate embedding columns within sampling tool by @Jeadie in https://github.com/spiceai/spiceai/pull/4722
- Validate primary key columns during accelerated dataset initialization by @sgrebnov in https://github.com/spiceai/spiceai/pull/4736

**Full Changelog**: https://github.com/spiceai/spiceai/compare/v1.0.2...v1.0.3

Resources​

Community​

Spice.ai started with the vision to make AI easy for developers. We are building Spice.ai in the open and with the community. Reach out on Discord or by email to get involved.

Spice.ai v0.6.1-alpha

· 2 min read
Luke Kim
Luke Kim
Founder and CEO of Spice AI

Announcing the release of Spice.ai v0.6.1-alpha! 🌶

Building upon the Apache Arrow support in v0.6-alpha, Spice.ai now includes new Apache Arrow data processor and Apache Arrow Flight data connector components! Together, these create a high-performance bulk-data transport directly into the Spice.ai ML engine. Coupled with big data systems from the Apache Arrow ecosystem like Hive, Drill, Spark, Snowflake, and BigQuery, it's now easier than ever to combine big data with Spice.ai.

And we're also excited to announce the release of Spice.xyz! 🎉

Spice.xyz is data and AI infrastructure for web3. It’s web3 data made easy. Insanely fast and purpose designed for applications and ML.

Spice.xyz delivers data in Apache Arrow format, over high-performance Apache Arrow Flight APIs to your application, notebook, ML pipeline, and of course through these new data components, to the Spice.ai runtime.

Read the announcement post at blog.spice.ai.

Spice.xyz

New in this release​

Now built with Go 1.18.

Dependency updates​

  • Updates to React 18
  • Updates to CRA 5
  • Updates to Glide DataGrid 4
  • Updates to SWR 1.2
  • Updates to TypeScript 4.6

Resources​

Community​

Spice.ai started with the vision to make AI easy for developers. We are building Spice.ai in the open and with the community. Reach out on Discord or by email to get involved. We will also be starting a community call series soon!

Spice.ai v0.6-alpha

· 3 min read
Phillip LeBlanc
Phillip LeBlanc
Co-Founder and CTO of Spice AI

Announcing the release of Spice.ai v0.6-alpha! 🏹

Spice.ai now scales to datasets 10-100 larger enabling new classes of uses cases and applications! 🚀 We've completely rebuilt Spice.ai's data processing and transport upon Apache Arrow, a high-performance platform that uses an in-memory columnar format. Spice.ai joins other major projects including Apache Spark, pandas, and InfluxDB in being powered by Apache Arrow. This also paves the way for high-performance data connections to the Spice.ai runtime using Apache Arrow Flight and import/export of data using Apache Parquet. We're incredibly excited about the potential this architecture has for building intelligent applications on top of a high-performance transport between application data sources the Spice.ai AI engine.

Apache Arrow

Highlights in v0.6-alpha​

Massive improvement in data loading performance and dataset scale​

From data connectors, to REST API, to AI engine, we've now rebuilt Spice.ai's data processing and transport on the Apache Arrow project. Specifically, using the Apache Arrow for Go implementation. Many thanks to Matt Topol for his contributions to the project and guidance on using it.

This release includes a change to the Spice.ai runtime to AI Engine transport from sending text CSV over gGPC to Apache Arrow Records over IPC (Unix sockets).

This is a breaking change to the Data Processor interface, as it now uses arrow.Record instead of Observation.

Benchmarking v0.6​

Performance Graph

Before v0.6, Spice.ai would not scale into the 100s of 1000s of rows.

FormatRow NumberData SizeProcess TimeLoad TimeTransport timeMemory Usage
csv2,000163.15KiB3.0005s0.0000s0.0100s423.754MiB
csv20,0001.61MiB2.9765s0.0000s0.0938s479.644MiB
csv200,00016.31MiB0.2778s0.0000sNA (error)0.000MiB
csv2,000,000164.97MiB0.2573s0.0050sNA (error)0.000MiB
json2,000301.79KiB3.0261s0.0000s0.0282s422.135MiB
json20,0002.97MiB2.9020s0.0000s0.2541s459.138MiB
json200,00029.85MiB0.2782s0.0010sNA (error)0.000MiB
json2,000,000300.39MiB0.3353s0.0080sNA (error)0.000MiB

After building on Arrow, Spice.ai now easily scales beyond millions of rows.

FormatRow NumberData SizeProcess TimeLoad TimeTransport timeMemory Usage
csv2,000163.14KiB2.8281s0.0000s0.0194s439.580MiB
csv20,0001.61MiB2.7297s0.0000s0.0658s461.836MiB
csv200,00016.30MiB2.8072s0.0020s0.4830s639.763MiB
csv2,000,000164.97MiB2.8707s0.0400s4.2680s1897.738MiB
json2,000301.80KiB2.7275s0.0000s0.0367s436.238MiB
json20,0002.97MiB2.8284s0.0000s0.2334s473.550MiB
json200,00029.85MiB2.8862s0.0100s1.7725s824.089MiB
json2,000,000300.39MiB2.7437s0.0920s16.5743s4044.118MiB

New in this release​

Resources​

Community​

Spice.ai started with the vision to make AI easy for developers. We are building Spice.ai in the open and with the community. Reach out on Discord or by email to get involved. We will also be starting a community call series soon!