chdb-serverless
Run a chDB (in-process ClickHouse) analyst on serverless — one app and one image across AWS Lambda, Google Cloud Run, and Azure Container Apps, with two plug seams that keep the core unchanged as you grow:
- store seam —
local:(this package, L1) →durable:(S3-backed object, L2) →memory:(analytical agent memory, L3), chosen byCHDB_STORE; - model seam — Anthropic / OpenAI / any OpenAI-compatible server, chosen by
CHDB_MODEL, so the analyst is not anchored to one LLM.
This repo is the base. The walkthroughs — why in-process, the cold-start comparison across clouds, the tier ladder — live in the chDB cookbook serverless series. Heritage: this repo pioneered chDB-on-Lambda in 2023 (thanks @lmangani); the maintained, multi-cloud version is here.
Use as a template (clone and deploy)
git clone https://github.com/chdb-io/chdb-lambda
cd chdb-lambda
export ANTHROPIC_API_KEY=sk-... # optional — omit for /query only
deploy/aws-lambda/deploy.sh # or gcp-cloud-run/ or azure-container-apps/
Auth posture (read before deploying): /query runs caller-supplied
ClickHouse SQL and /ask spends model tokens, so every lane deploys
non-public by default — AWS Lambda uses an AWS_IAM Function URL, and
Cloud Run / Container Apps use private ingress. Set PUBLIC=1 to opt into an
unauthenticated endpoint for a throwaway demo, and put your own auth in front
before exposing anything real.
Use as a package (import, don't copy)
pip install chdb-serverless[anthropic] # or [openai]
from chdb_serverless import analyst_app, open_store
app = analyst_app() # FastAPI: /health /query /ask
store = open_store() # CHDB_STORE selects the tier
Seams
| Env | Values | Meaning |
|---|---|---|
CHDB_STORE |
local:/path · durable:s3://… · memory:s3://… |
which tier (L1 now; L2/L3 via extras) |
CHDB_MODEL |
anthropic:claude-opus-4-8 · openai:gpt-4.1 · openai:llama3@http://host/v1 |
which LLM (any OpenAI-compatible server via @base_url) |
Extras: [anthropic] [openai] (LLM providers), [durable] (L2), [memory] (L3).
Layout
src/chdb_serverless/ app: server.py (/health /query /ask), agent.py, store.py, models/
Dockerfile one image; bakes the dataset; carries the Lambda Web Adapter (inert off Lambda)
deploy/ per-cloud deploy.sh + teardown.sh (aws-lambda / gcp-cloud-run / azure-container-apps)
Metadata
Release files for chdb-serverless 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| chdb_serverless-0.1.0.tar.gz | 24.1 kB | Without provenance | Details for chdb_serverless-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| chdb_serverless-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for chdb_serverless-0.1.0-py3-none-any.whl |
Total release size: 43.0 kB
Release files / chdb_serverless-0.1.0.tar.gz
| Download URL | chdb_serverless-0.1.0.tar.gz |
|---|---|
| Size | 24.1 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
19b242a01ef59a3d58f029da3ab1dbc6ac46f03f64a22202ca941ab2f9f83a8b
|
|
BLAKE2b-256 checksum How to use checksums |
a5d8a185af71e802a49e7abb1c99f9b416968ffb95291189d34d6e700489e23b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.14.6
|
Release files / chdb_serverless-0.1.0-py3-none-any.whl
| Download URL | chdb_serverless-0.1.0-py3-none-any.whl |
|---|---|
| Size | 18.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
116ac850dc8e41ebe28ec47c06241b07b90834f02cef1a94817afcbef4166378
|
|
BLAKE2b-256 checksum How to use checksums |
fe6fa730eab88bb5a8754c025c045c89f902a225b4a1161b8cc36450c40066a2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.14.6
|
Security for chdb-serverless 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.