Import Data via Console

This guide walks you through importing data from object storage (AWS S3, Google Cloud Storage, Azure Blob Storage) into VeloDB using the Console visual interface.

Prerequisites

Before starting, ensure you have:

• Object storage bucket with data files
• Access credentials (Access Key ID and Secret Access Key)
• VeloDB Cloud account with an active cluster (see Quick Start)

Try it with Sample Data

Use these credentials to try the S3 import with our sample dataset:

Field	Value
AK	AKIA3AUKURBS74337SNB
SK	ygbR1HGNvMZDTo4DNUWJx0mblpMTF+QpBCCBfxFF
Object Storage Path	https://velodb-import-data-us-east-1.s3.us-east-1.amazonaws.com/ssb-flat-sf1/*.parquet

This sample dataset contains SSB (Star Schema Benchmark) data - a widely used benchmark for analytical databases. The dataset includes ~6 million rows of denormalized sales data with 42 columns covering orders, customers, suppliers, and products.

SSB Dataset Schema (42 columns)

Column Group	Columns
Order	lo_orderkey, lo_linenumber, lo_orderdate, lo_commitdate, lo_orderpriority, lo_shippriority, lo_shipmode, lo_year, lo_month, lo_weeknum
Metrics	lo_quantity, lo_extendedprice, lo_discount, lo_revenue, lo_supplycost, lo_tax
Date	d_datekey, d_dayofweek, d_month, d_yearmonth
Customer	c_custkey, c_name, c_nation, c_region, c_city, c_mktsegment
Supplier	s_suppkey, s_name, s_nation, s_region, s_city
Product	p_partkey, p_name, p_brand, p_category, p_mfgr, p_color, p_type, p_size, p_container

tip

This is read-only sample data. You can use it to follow along with the tutorial steps below.

Step 1: Connection

Navigate to Data > Import in the VeloDB Console sidebar, then click Create new and select Object Storage S3.

Configuration

Field	Description
Task Name	A unique name for this import task (e.g., sales_data, user_logs)
Comment	(optional) Description of the import task
Authentication	Select Access Key authentication
AK	Your Access Key ID (e.g., AKIAIOSFODNN7EXAMPLE)
SK	Your Secret Access Key
Object Storage Path	URL to your data (see format below)

Object Storage Path Format

https://<bucket-name>.s3.<region>.amazonaws.com/<path>/<filename>

Example path formats (replace with your actual bucket and file paths):

• Single file: https://my-bucket.s3.us-west-1.amazonaws.com/data/orders.csv
• Multiple files with wildcard: https://my-bucket.s3.us-west-1.amazonaws.com/data/*.csv
• Parquet files: https://my-bucket.s3.us-west-1.amazonaws.com/warehouse/*.parquet

warning

The object storage bucket must be in the same region as your VeloDB cluster for optimal performance.

Click Next to proceed.

Step 2: Incoming Data

Configure how VeloDB should parse your data files.

File Configuration

Field	Description
File Type	Select your file format: CSV, Parquet, ORC, or JSON
File Compression	Auto-detect or specify: GZ, BZ2, LZ4, LZO, DEFLATE, ZSTD, ZLIB
Specify Delimiter	Column separator (, for CSV, \t for TSV)
Enclose	Quote character for text fields (usually leave empty)
Escape	Escape character (usually leave empty)
Trim Double Quotes	Whether to trim quotes from values
File Size	Set a size limit or leave as Unlimited

Loading Configuration

Field	Description
Strict Mode	ON = reject rows with errors, OFF = skip bad rows

info

For standard CSV files, the default settings usually work. Leave Enclose and Escape empty to avoid parsing errors.

Click Next to proceed.

Step 3: Configure Table

Preview your data and configure the destination table.

Data Preview

The console displays a preview of your data with:

• Auto-detected column names (c1, c2, c3, ... or from CSV header row)
• Sample data rows
• Inferred column types

Load Data To

Option	Description
New Table	Create a new table with auto-generated schema
Existing Table	Load into an existing VeloDB table

Table Configuration

Field	Description
Database	Choose target database from dropdown
Table Name	Name for the new table (e.g., orders, user_events, products)

Step 4: Advanced Settings

Configure table model and distribution settings.

Table Models

Model	Use Case	Example
DUPLICATE	Raw data, ad-hoc queries - retains all rows as written	Event logs, clickstream, raw transactions
UNIQUE	Data with updates - keeps only latest row per key	User profiles, product catalog, dimension tables
AGGREGATE	Pre-aggregated metrics - auto-aggregates by key columns	Sales summaries, time-series metrics, counters

Distribution Settings

Field	Description
Sorting Key	Column(s) for data ordering - choose columns frequently used in WHERE clauses or JOINs
Partition	Enable for partitioning by date/time or other dimensions to improve query performance
Bucket Key	Column for hash distribution across nodes (use tenant_id or user_id to improve query concurrency)
Bucket Number	Number of data buckets (AUTO is recommended for most cases)
Properties	Additional table properties (usually leave empty)

Click Next to proceed, then Submit to start the import.

Step 5: Monitor Import

After submitting, you'll see your import task in the Import list.

Verify Import

After the import completes, verify your data in the SQL Editor:

-- Check row count
SELECT COUNT(*) FROM your_database.your_table;

-- Preview data
SELECT * FROM your_database.your_table LIMIT 10;

-- Check table schema
DESC your_database.your_table;

Sample Queries for SSB Data

If you imported the sample SSB dataset, try these analytical queries:

-- Total revenue by year
SELECT
    lo_year,
    SUM(lo_revenue) as total_revenue
FROM ssb_flat
GROUP BY lo_year
ORDER BY lo_year;

-- Top 10 customers by revenue
SELECT
    c_name,
    c_nation,
    SUM(lo_revenue) as total_revenue
FROM ssb_flat
GROUP BY c_name, c_nation
ORDER BY total_revenue DESC
LIMIT 10;

-- Revenue by region and year
SELECT
    c_region,
    lo_year,
    SUM(lo_revenue) as revenue,
    COUNT(*) as order_count
FROM ssb_flat
GROUP BY c_region, lo_year
ORDER BY c_region, lo_year;

-- Product category performance
SELECT
    p_category,
    p_brand,
    SUM(lo_revenue) as revenue,
    AVG(lo_discount) as avg_discount
FROM ssb_flat
GROUP BY p_category, p_brand
ORDER BY revenue DESC
LIMIT 20;

Troubleshooting

Issue	Solution
"Can not found files"	Check object storage path format and trailing slash
"Access Denied"	Verify AK/SK credentials and IAM permissions
Connection timeout	Ensure bucket is in same region as VeloDB
Parsing errors	Leave Enclose and Escape fields empty
Wrong column types	Use Existing Table with predefined schema