dbs3-pycurl

3.18.3 verified Fri May 01 auth: no python

DBS3 Pycurl Client is a Python library providing a pycurl-based HTTP client for interacting with the CMS Data Bookkeeping Service (DBS). Version 3.18.3 is currently available, with last release in 2023-2024, and it follows the DBS3 release schedule.

pip install dbs3-pycurl

Common errors

error ImportError: No module named dbs.apis.dbsClient ↓

cause The package dbs3-pycurl is not installed or installed in a different environment.

fix

Run 'pip install dbs3-pycurl' and ensure it is in the same Python environment.

error pycurl.error: (7, 'Failed to connect to address') ↓

cause Network issue or incorrect DBS server URL.

fix

Check the URL for typos and ensure network connectivity (e.g., VPN or CERN proxy).

error dbs.exceptions.DBSException: Invalid input parameter ↓

cause The dataset name or path is malformed or does not exist.

fix

Verify the dataset name format (e.g., /primary/proc/datatype). Use listDatasets to check availability.

Warnings

gotcha The DBSClient and DBSReader classes are not thread-safe. Do not share instances across threads without locks. ↓

fix Create a new client instance per thread or use threading locks.

deprecated The 'dbs.apis.dbsClient' API is gradually being replaced by REST-based clients. Check for 'DBS3Client' or direct REST calls in newer releases. ↓

fix Monitor release notes for deprecation timelines and migrate to REST.

gotcha The library uses pycurl under the hood; ensure that the system has libcurl installed and pycurl is compiled with SSL support to avoid connection errors. ↓

fix Install pycurl with SSL support, e.g., 'PYCURL_SSL_LIBRARY=openssl pip install pycurl'.

Imports

DBSClient
```
from dbs.apis.dbsClient import DBSClient
```
Common mistake: trying to import DBSClient directly from dbs or dbs3
DBSReader
```
from dbs.apis.dbsClient import DBSReader
```
DBSReader is used for read-only access
DbInfo
```
from dbs.utils.dbinfo import DbInfo
```
Utility class for database info

Quickstart

Initialize DBSReader with production URL and fetch dataset summary.

from dbs.apis.dbsClient import DBSReader
import os

# Example: fetch dataset summary by dataset
reader = DBSReader(url="https://cmsweb.cern.ch/dbs/prod/global/DBSReader")
dataset = "/Sample/Sample-4342-0003/GEN-SIM"
summary = reader.listDatasetSummary(dataset=dataset)
print(summary)