Connecting to CephRDS (S3 Object Storage)

Researchers, PIs, Students

Overview

CephRDS is Research Computing's on-premises research data storage, built on Ceph. Labs reach their buckets through an S3-compatible interface only, so you use an S3 client rather than mapping a network drive. Other access methods, such as NFS, are not offered at this time.

This article covers how to request storage and keys, and how to connect to the CephRDS endpoint (https://rds.ucr.edu) with Cyberduck (graphical), rclone (command line) and Python.

Before you request storage, note:

Pilot service. CephRDS is in pilot. Capacity is limited, terms may change, and each request is reviewed individually. See CephRDS.
Network. CephRDS is reachable from the campus network only. Off campus, connect to the UCR campus VPN (Cisco Secure Client) first.
Data allowed. P1 and P2 data only. Do not store P3 or P4 data on CephRDS.
Costs. Pilot rates are $50 per TB per year (CephRDS Sustainability and Financial Model (Research Computing, Sep 2026), as of Sep 2026) for rented capacity, or $250 per usable TB, one time, for the 5-year hardware lifecycle for purchased capacity. The terms set out when your storage is allocated are the ones that apply. See Costs.

Requesting storage and access keys

Send a request. The PI or lab lead contacts Research Computing through the UCR Support Portal or by email to research-computing@ucr.edu.
Include these details:
- PI or lab lead name and NetID
- Department and college
- Estimated initial capacity (for example, 10 TB)
- Lab members (NetIDs) who need access to the bucket
- The kind of data you plan to store, so its data classification can be confirmed
Review and setup. Research Computing reviews the request and works with ITS to create the storage. When it is set up, the S3 Access Key ID and Secret Access Key are sent to the PI through a secure channel.

Keep your keys private. Do not email them, paste them into chat, or commit them to a code repository. If a key may have been exposed, contact research-computing@ucr.edu so it can be replaced.

1. Using Cyberduck (graphical)

Cyberduck is an open-source graphical client for macOS and Windows, suited to drag-and-drop transfers. KB019 covers it and a Linux alternative in more detail.

Download and install Cyberduck from https://cyberduck.io.
Open a new connection:
- Click Open Connection.
- From the drop-down at the top, select Amazon S3.
Enter the connection details:
- Server: rds.ucr.edu
- Port: 443
- Access Key ID: your access key
- Secret Access Key: your secret key
Click Connect. Your buckets are listed, and you can drag and drop files.

2. Using rclone (command line)

rclone is a command-line tool for cloud and object storage, suited to large transfers and scripted copies. It runs on Linux, macOS and Windows.

Configuration

Start the configuration tool:
```
rclone config
```
Answer the prompts to create a new remote:
- n/s/q>: n (new remote)
- name>: a name for the remote, for example ucr-ceph
- Storage>: s3 (Amazon S3 Compliant Storage Providers)
- provider>: Ceph (Ceph Object Storage). Do not select "Amazon S3".
- env_auth>: false (enter credentials in the next step)
- access_key_id>: your Access Key ID
- secret_access_key>: your Secret Access Key
- region>: leave blank (press Enter)
- endpoint>: https://rds.ucr.edu
- location_constraint>: leave blank (press Enter)
- acl>: leave blank (press Enter)
- Accept the defaults for the remaining prompts, then save and quit.

Basic commands

Replace my-lab-bucket with your bucket name.

List buckets:
```
rclone lsd ucr-ceph:
```
List files in a bucket:
```
rclone ls ucr-ceph:my-lab-bucket
```

Copy a local folder to CephRDS, with progress shown:

rclone copy /path/to/local/data/ ucr-ceph:my-lab-bucket/folder/ -P

To use a bucket like a local folder, see KB020: Mounting CephRDS buckets as local drives.

3. Using Python (boto3)

Use the AWS SDK for Python (boto3) with two settings: point endpoint_url at CephRDS, and use path-style addressing. You do not need a region.

import os
import boto3
from botocore.client import Config

s3 = boto3.client(
    "s3",
    endpoint_url="https://rds.ucr.edu",
    aws_access_key_id=os.environ["CEPHRDS_ACCESS_KEY"],
    aws_secret_access_key=os.environ["CEPHRDS_SECRET_KEY"],
    config=Config(s3={"addressing_style": "path"}),
)

for obj in s3.list_objects_v2(Bucket="my-lab-bucket").get("Contents", []):
    print(obj["Key"], obj["Size"])

KB018: Accessing CephRDS with Python (boto3) has a fuller example, including uploads and downloads.

Getting help

Contact Research Computing at research-computing@ucr.edu.

Owner: Research Computing Reviewed: 4 Oct 2026