What are the storage devices used in networking

Networked storage exists so that data isn’t trapped on a single server’s local disks — it can be shared, protected, and scaled independently of any one machine. This article surveys the storage devices and systems commonly deployed in networked environments, how they fit together, and practical guidance on setting up and managing each.

The Networked Storage Landscape

graph TD
    Client[Clients/Servers] --> NAS[NAS: File-Level Storage]
    Client --> SAN[SAN: Block-Level Storage]
    Client --> Object[Object Storage: API-based]
    SAN --> Array[Storage Array Controllers]
    NAS --> Array2[NAS Head + Disk Shelves]
    Object --> Cluster[Object Storage Cluster Nodes]

NAS (Network Attached Storage)

A NAS device is a self-contained system connected to the LAN that serves files over NFS (Linux/Unix) or SMB/CIFS (Windows), appearing to clients as ordinary shared folders.

Typical NAS Hardware Range

Setting Up NFS Sharing on a Linux NAS

sudo apt install nfs-kernel-server
sudo mkdir -p /export/data
sudo chown nobody:nogroup /export/data

echo "/export/data 10.0.0.0/24(rw,sync,no_subtree_check)" | sudo tee -a /etc/exports
sudo exportfs -ra
sudo systemctl restart nfs-kernel-server

sudo apt install nfs-kernel-server
sudo mkdir -p /export/data
sudo chown nobody:nogroup /export/data

echo "/export/data 10.0.0.0/24(rw,sync,no_subtree_check)" | sudo tee -a /etc/exports
sudo exportfs -ra
sudo systemctl restart nfs-kernel-server

Client-side mount:

sudo mount -t nfs 10.0.0.60:/export/data /mnt/data

sudo mount -t nfs 10.0.0.60:/export/data /mnt/data

Setting Up Samba (SMB) Sharing on Linux

sudo apt install samba

sudo tee -a /etc/samba/smb.conf << 'EOF'

sudo apt install samba

sudo tee -a /etc/samba/smb.conf << 'EOF'

path = /srv/samba/shared browsable = yes writable = yes guest ok = no valid users = @sambashare EOF sudo systemctl restart smbd sudo smbpasswd -a myuser

SAN (Storage Area Network) Devices

A SAN presents block-level storage over a dedicated network (Fibre Channel or iSCSI), with each connected server treating the storage as if it were a locally attached disk. Covered in depth in the companion Storage Array and Storage Interfaces/Protocols articles — the key SAN-specific devices are:

• Storage array controllers: Manage RAID, caching, and host connectivity.
• FC/iSCSI switches: Build the fabric or network connecting hosts to the array.
• HBAs/iSCSI initiators: The host-side connection point into the SAN.

Tape Storage (Still Relevant for Networked Backup)

Despite being decades old, LTO (Linear Tape-Open) tape remains widely used for long-term archival and offsite backup, due to its low cost per terabyte and strong resilience for cold, rarely-accessed data.

Tape libraries are typically connected via SAS or Fibre Channel and integrated with backup software that manages tape rotation, cataloging, and offsite handling.

Object Storage Nodes

Object storage systems (covered in the Storage Array article) run across clusters of standard servers, each contributing disks to a shared, distributed pool. Data is spread and protected using erasure coding rather than traditional RAID, allowing the system to tolerate multiple node or drive failures across a large cluster.

Example: Interacting with S3-Compatible Object Storage (Python)

import boto3

s3 = boto3.client(
    "s3",
    endpoint_url="http://10.0.0.70:9000",
    aws_access_key_id="myaccesskey",
    aws_secret_access_key="mysecretkey"
)

# Upload a file
s3.upload_file("backup.tar.gz", "my-bucket", "backups/backup.tar.gz")

# List objects in a bucket
response = s3.list_objects_v2(Bucket="my-bucket")
for obj in response.get("Contents", []):
    print(obj["Key"], obj["Size"])

import boto3

s3 = boto3.client(
    "s3",
    endpoint_url="http://10.0.0.70:9000",
    aws_access_key_id="myaccesskey",
    aws_secret_access_key="mysecretkey"
)

# Upload a file
s3.upload_file("backup.tar.gz", "my-bucket", "backups/backup.tar.gz")

# List objects in a bucket
response = s3.list_objects_v2(Bucket="my-bucket")
for obj in response.get("Contents", []):
    print(obj["Key"], obj["Size"])

This example works against any S3-compatible object store (AWS S3, MinIO, Ceph RGW, etc.), illustrating the API-driven nature of object storage compared to file/block access.

Storage Devices in a Data Center Network Context

graph TD
    subgraph Data Center
    Compute[Server Cluster] --- SANSwitch[SAN Switch Fabric]
    SANSwitch --- Array[Block Storage Array]
    Compute --- LAN[Ethernet LAN]
    LAN --- NASHead[NAS Head]
    NASHead --- Shelf[Disk Shelves]
    LAN --- ObjCluster[Object Storage Cluster]
    Array --- Backup[Backup Target / Tape Library]
    end

Redundancy and High Availability in Networked Storage

Snapshots and replication protect against different failure modes than RAID — RAID protects against a drive failure, but not against someone accidentally deleting a file, which is where snapshots come in.

Practical Example: Verifying Storage Connectivity from a Server

# For iSCSI-attached storage
sudo iscsiadm -m session -P 3

# For NFS mounts
mount | grep nfs
showmount -e 10.0.0.60

# For general block device visibility
lsblk
multipath -ll

# For iSCSI-attached storage
sudo iscsiadm -m session -P 3

# For NFS mounts
mount | grep nfs
showmount -e 10.0.0.60

# For general block device visibility
lsblk
multipath -ll

Cisco Example: Verifying SAN Port Connectivity for Storage

Switch# show interface fc1/1
Switch# show flogi database
Switch# show fcns database

Switch# show interface fc1/1
Switch# show flogi database
Switch# show fcns database

Best Practices

• Match storage device type to workload: NAS for shared files, SAN for databases/VMs, object storage for backup/archive/unstructured scale.
• Always plan redundancy at every layer — controller, network path, and drive — not just one.
• Separate storage traffic (iSCSI, NFS) onto dedicated VLANs or physical networks where performance matters.
• Use snapshots in addition to, not instead of, real backups — a snapshot on the same array doesn’t protect against array-level failure.
• Regularly test backup/restore procedures, not just backup completion — an untested backup is not a reliable backup.
• Monitor capacity, performance, and error counters proactively across all networked storage devices, not just when something breaks.

Troubleshooting

Further Reading

• SNIA Storage Networking Industry Association Education
• RFC 7530 — NFS Version 4 Protocol
• Samba Project Documentation
• LTO Ultrium Technology Overview
• MinIO S3-Compatible Object Storage Documentation

Conclusion

Networked storage spans a wide spectrum — from simple file-sharing NAS boxes to high-performance SAN fabrics, tape archives, and massively scalable object storage clusters. Understanding which device type fits which workload, and building redundancy in at every layer, is what separates a resilient storage environment from one that’s one failure away from a serious outage.