Storage Solution

Trainer: Dipen Khadka

B.E Computer NCIT
CCNA, JNCIA
Cypher Technology P. Ltd.
dipen@cyphertechnepal.com

Trainer: Santosh Parajuli

BCIS Apex
VCP, DEA-Associate, NSE
Cypher Technology P. Ltd.
santosh@cyphertechnepal.com
 Agenda:
Session: • Storage Concept
• Type of Storage
Day 3
• RAID Level
Session 1: 12:00 – 1:00 • Data Storage Technologies( DAS, NAS, SAN)
Session 2: 1:10 - 2:00 • Capacity Planning, Storage Expansion
Session 3 : 2:45 – 4:15 • Practical
• Configuration and Managing a NAS Devices.
• Setting Up a SAN Environment with LUNs.
 what is Storage?
• Storage refers to the methods and technologies used to retain
digital data.
• It is a core function and fundamental component of computers
• Storage is required for following reason:
➢ The main memory is temporary memory. The storage is
required to store data and programs permanently.
➢The capacity is limited. The storage is required to store a large
amount of data and program.
 Basic Unit of Storage
• Bit: Computer works with binary digit. These digits are in the form of 0’s and 1’s. A
binary digit is called bit. One bit takes one storage location in memory. It is the
smallest unit for data storage.
• Byte: A collection of eight bits is called bytes. It is used to store single character. The
capacity of the memory of the storage is expressed in terms of bytes.
• Kilobyte : A kilobyte consists of 1024 bytes. It is denoted by KB.
• Megabyte: A megabyte consists of 1024 kilobytes. It is denoted by MB.
• Gigabyte: A gigabyte consists of 1024 megabytes. IT is denoted by GB.
• Terabyte: A terabyte consists of 1024 gigabytes. It is denoted by TB.
Data Storage Throughout History
Type of Storage
• Object Storage
Stores data as objects, each with metadata and a unique identifier.
Accessed via APIs (e.g., HTTP/HTTPS).
• Examples: Amazon S3, Google Cloud Storage.
• Use Cases: Unstructured data, backups, big data analytics.
• Advantages: Unlimited scalability, cost-effective for massive datasets.

• Block Storage
Data is split into fixed-sized blocks and managed by the OS or applications.
High performance and low latency.
• Examples: Amazon Elastic Block Store (EBS), SANs.
• Use Cases: Databases, virtual machines, transactional systems.
• Advantages: High performance, flexibility.

• File Storage
Organized in a hierarchical structure (files and folders).
Accessed via protocols like NFS or SMB.
• Examples: NAS devices, Amazon EFS.
• Use Cases: Shared drives, collaboration, content management.
• Advantages: User-friendly, network sharing.
 Type of Storage
• Based on Media Type
• Magnetic Storage
Uses magnetic properties to store data.
Examples: HDDs, magnetic tapes.
Use Cases: Long-term storage, cost-effective solutions.
Advantages: Cost-efficient, large capacity.
• Flash Storage
Non-volatile storage using NAND flash memory.
Examples: SSDs, NVMe drives.
Use Cases: High-performance computing, consumer devices.
Advantages: Low latency, high speed.
 Type of Storage
Feature HDD SSD NVMe
use spinning magnetic SSDs use NAND flash NVMe is a protocol designed
disks (platters) and a memory to store data specifically for SSDs,
mechanical arm (actuator) electronically, with no leveraging the PCIe interface
to read and write data. moving parts. for faster data transfer.
Speed 50–150 MB/s 200-3500 MB/s Up to 7,000 MB/s (PCIe 4.0)
Capacity 500GB–20TB 128 GB – 8 TB 256GB–8TB
Interface SATA, SAS SATA, PCIe PCIe (x2, x4, x8 lanes)
Use Case Archiving, bulk storage General computing, gaming Professional, high-
performance
RAID
Redundant Array
of Independent Disks
 What is RAID?

• Stands for Redundant Array of Independent Disks.

• It's a technology that enables greater levels of performance, reliability and/or large
volumes when dealing with data.
• How ?? By concurrent use of two or more 'hard disk drives’.
• How Exactly ?? Mirroring, Stripping (of data) and Error correction techniques
combined with multiple disk arrays give you the reliability and performance.
 RAID Advantages
• Fault Tolerance.
• Higher Data Security.
• Increase the parity check for any possible system crash.
• Reading and Writing of data done at simultaneously.
• Improved Availability and performance.
• Ensures data reliability.
 Types of RAID
• Commonly used ones:
• RAID 0
• RAID 1
• RAID 5
• RAID 10
• Other types used ... but rarely: RAID 2,3,4,6,50...
 RAID 0
• RAID 0 is a storage technology designed to improve
performance by distributing data across multiple drives
in a process called striping.
• Key Features:
• Data Striping: Data is split into smaller chunks(stripes)
and written across all available drives.
• No Redundancy: RAID 0 does store duplicate data
or parity information.
• Higher Performance: Improve read and write speeds
• because multiple drives work simultaneously to handle
data.
• Requires at least two drives to implement.
 RAID 0
• Example of RAID 0 in Action
• Scenario:
• Two drives of 500GB each in a RAID 0 setup.
• Data chunks are striped alternately across the two drives.
• Benefits:
• Total capacity: 1TB.
• Read/write speeds effectively double because both drives operate simultaneously.
• Risk:
• If either drive fails, all data in the RAID 0 array is lost.
Data Integrity
 RAID 1

• RAID 1 focuses on data redundancy by creating exact copy

(mirrors) of the data on two or more drives.
Key Features:
• Mirroring: Data is written identically to two( or more)
drives.
• High Redundancy: Ensures data availability even in the
even of a single drive failure.
• Read Performance Improvement: Some RAID controllers
can use both drives to serve read requests simultaneously,
increasing read speeds.
• Requires at least two drives.
 RAID 1
• Example of RAID 1 in Action
• Scenario:
• Two drives of 2TB each in a RAID 1 setup.
• Data written to Drive 1 is mirrored on Drive 2.
• Benefits:
• Total capacity: 4TB, but only 2TB is usable.
• If Drive 1 fails, Drive 2 has an exact copy of the data.
• Risk :
• Redundancy protects against single drive failure but does not protect against
simultaneous drive failures.
Data Integrity
 RAID 5
• Striping with Parity
• It combines performance,
redundancy and efficient use of
storage by distributing data
and parity( error-checking
information) across multiple
drives.
• It is one of the most popular
RAID Level.
• Key Features: RAID 5
• Data is striped across all drives in the
array, with parity information
distributed evenly.
• Can tolerate the failure of one drive
with out data loss.
• Offer a good balance of storage
capacity, performance and fault
tolerance.
• Requires a minimum of three drives
to implement.
Disadvantage:
RAID 5
• If a drive fails, rebuilding the array can take a long time and impacts performance.
• If another drive fails during the rebuild process, all data is lost.
Example of RAID 5 in Action
• Scenario:
Four drives of 2TB each in a RAID 5 setup.
Data is striped, and parity is distributed across all drives.
• Benefits:
Total capacity: 6TB (1 drive's capacity is used for parity).
If any one drive fails, the data can be rebuilt using parity information.
• Risk:
Protects against a single drive failure but not against dual drive failures
Data Integrity
RAID 5
Cost vs Capacity

1 TB 1 TB 1 TB 1 TB

Raw Capacity: 4 TB
Usable Capacity: N-1 = 4 -1 = 3 TB
 RAID 10
• Mirroring and Striping
• RAID 10, also known as RAID 1+0, combines the features of RAID1 ( mirroring) and RAID 0
(striping).
• Key Feature:
• Data is mirrored across pairs of drives( RAID 1), ensuring redundancy.
• Mirrored pairs are striped ( RAID 0) for improved performance.
• Offers fast read and write speeds due to striping.
• Can tolerate multiple drive failures as long as no mirrored pair is lost entrirely.
• Require at least four drives.
 RAID 10
• RAID 10 Configuration
• Mirrored Pairs:
Disk 1 is mirrored with Disk 2.
Disk 3 is mirrored with Disk 4.
• Striping:
Data is striped across the two mirrored pairs for better read/write performance.
• Available Storage Capacity
Each mirrored pair provides the storage of one drive (2 TB).
Two mirrored pairs combined provide 2 TB + 2 TB = 4 TB, but only half of the total raw capacity is
usable due to mirroring.
• Total Usable Storage in RAID 10 with 4 Drives: 4 TB.
Data Integrity
Data Integrity
 What's happening in present day?

RAID 6:
• a. It is seen as the best way to
guarantee data integrity as it uses
double parity.
• b. Require at least 4 drive.
• c. It has a drawback though of longer
write time.
Data Integrity
 Quiz
You have 8 drives, each with 1 TB capacity, in a RAID 10 configuration. What is the total usable capacity?

4 TB
In a RAID 6 array with 8 drives, how many drives can fail without data loss?

2
"A police department maintains a criminal records database that requires moderate performance, redundancy,
and cost-efficiency. Which RAID level do you recommend?"
RAID 5
"A government IT department is setting up a disaster recovery site for its public-facing websites. The storage must
ensure redundancy and quick recovery times. Which RAID level do you recommend?"

RAID 10
"A research lab runs machine learning models requiring fast access to massive datasets. Redundancy is not
critical, as the data can be regenerated if lost. Which RAID level do you recommend?"

RAID 0
 RAID Implementation

Software based RAID:

• Software implementations are provided by many Operating Systems.
• A software layer sits above the disk device drivers and provides an abstraction
layer between the logical drives(RAIDs) and physical drives.
• Server's processor is used to run the RAID software.
• Used for simpler configurations like RAID O and RAID 1. .
 RAID Implementation

Hardware based RAID:

• Hardware RAID uses a dedicated RAID
controller, either built into the motherboard or
as a separate card, to manage the RAID array
independently of the operating system.
• Processor is not used for RAID calculations as
a separate controller present.
 Data Storage Technology
DAS( Direct-Attached Storage)
• Traditional way of implementing storage
• Storage is managed by a single host
• The storage device is directly connected to the computer or server via interfaces
like SATA (Serial ATA), SAS (Serial Attached SCSI), or USB.
• DAS does not require a network connection to the computer. It is a point-to-
point connection, where the storage is dedicated to a single system.
• Examples: Internal hard drives, external USB drives.
• Use Cases: Local storage for individual systems.
• Advantages: Simple, cost-effective.
 Data Storage Technology
NAS( Network Area Storage)
• NAS is a type of data storage system that connects to a network, allowing
multiple devices to access and share data across the network.
• NAS provides centralized storage that multiple users or devices can access
simultaneously.
• It’s typically used to store files, documents, videos, photos, backups, and other
data.
• NAS devices typically use file-level storage, meaning they manage files (rather
than blocks of data like SAN) and provide file access through network protocols
such as SMB (Server Message Block), NFS (Network File System).
 Data Storage Technology
NAS Architecture:
 Data Storage Technology

"What if we need block-level storage in a NAS environment?“

iSCSI to the Rescue!

 Data Storage Technology
• iSCSI (Internet Small Computer Systems Interface) enables a NAS to provide
block-level storage over an IP network.
• With iSCSI, a NAS device can act like a SAN (Storage Area Network), offering
block storage to clients and servers.
• How Does it Work?
• Logical Unit Numbers (LUNs) are created on the NAS to represent block storage
volumes.
• Initiators (client systems) access these LUNs over the network.
• The connected storage behaves like a directly attached hard drive for the client
system.
• Use case: Virtualization, Database, Server Application.
 Data Storage Technology
SAN( Storage Area Network)
• A Storage Area Network (SAN) is a high-speed, dedicated network that
connects storage devices to servers.
• SANs are designed to provide block-level data storage, offering fast, reliable, and
scalable solutions for data-intensive applications.
• Key Features of SAN:
• Block-Level Storage
• High Performance.
• Centralized Management:
 Data Storage Technology
Types of SAN Protocols:
• Fiber Channel (FC):
High-speed, dedicated storage network protocol.
Requires specialized hardware (Fiber Channel switches, HBAs).

• iSCSI:
Internet-based protocol using TCP/IP for SAN connectivity.
Cost-effective and uses existing Ethernet infrastructure.
 Data Storage Technology
SAN Architecture:
 Current SAN Technologies and Trends:
• Modern SAN Features:
• High-Speed Protocols
• NVMe-oF (NVMe over Fabrics):
• Uses NVMe protocol for ultra-low latency and high-speed storage access.
• Supported over Fiber Channel, RDMA, and TCP/IP networks.
• Ideal for AI/ML, real-time analytics, and other performance-critical applications.
• Fiber Channel (FC)
• Advanced Fiber Channel (16Gb, 32Gb, and 64Gb) provides high-speed, low-latency
connectivity.
• Ensures reliable communication in enterprise SAN environments.
• iSCSI (Internet Small Computer Systems Interface):
• Cost-effective block-level SAN protocol over Ethernet.
• Supports 10GbE, 25GbE, or 100GbE connections for faster data transfer
 IOPS
Definition:
IOPS measures the number of input/output operations a storage device can perform in a second. It
is a critical performance metric for assessing storage device speed, especially for random workloads.

Components Impacting IOPS:

• Read/Write Mix: Ratio of read to write operations.

• Random vs. Sequential Workloads: Random IOPS is generally lower than sequential IOPS due
to movement overhead in mechanical drives.
• Block Size: Smaller block sizes increase the number of IOPS but reduce throughput.
• Latency: Time taken for an operation, including seek time, rotational latency, and transfer time.
 IOPS based on Disk types
Spinning Disks (HDD):
IOPS Range:
• 7200 RPM HDD: 80–150 IOPS
• 10K RPM HDD: 120–200 IOPS
• 15K RPM HDD: 180–300 IOPS

SATA SSDs
IOPS Range:
• Read: 20,000–100,000 IOPS
• Write: 10,000–90,000 IOPS

NVMe SSDs
IOPS Range:
• Consumer-grade: 200,000–500,000 IOPS
• Enterprise-grade: 500,000–1,000,000+ IOPS
 IOPS based on Raid types
RAID 0 (Striping)
IOPS Impact:
• Read IOPS = Sum of all drives’ IOPS.
• Write IOPS = Sum of all drives’ IOPS.

RAID 1 (Mirroring)
IOPS Impact:
• Read IOPS = Sum of all drives’ IOPS (reads can be done in parallel).
• Write IOPS = Same as a single drive’s IOPS (writes must be mirrored).

RAID 5 (Striping with Parity)

IOPS Impact:
• Read IOPS = Sum of all drives’ IOPS.
• Write IOPS = (Number of drives - 1) × IOPS / 4 (due to parity calculations).
 IOPS based on Raid types
RAID 6 (Striping with Double Parity)
IOPS Impact:
• Read IOPS = Sum of all drives’ IOPS.
• Write IOPS = (Number of drives - 2) × IOPS / 6 (due to increased parity calculations).

RAID 10 (1+0, Striping and Mirroring)

IOPS Impact:
• Read IOPS = Sum of all drives’ IOPS.
• Write IOPS = Sum of mirrored pairs (equal to the number of drives divided by 2).
 Storage Capacity Planning and Expansion
1. Assess Current Storage Usage
• Inventory Check: List all storage devices and their capacities.
• Monitor Usage: Use monitoring tools (e.g., Zabbix, Grafana, or native tools) to track current
storage utilization.
• Data Categorization: Identify the types of data (e.g., operational, backup, archive) and their
growth patterns.
• Identify Bottlenecks: Look for signs of overutilized or underutilized storage.

2. Forecast Future Needs

• Growth Trends: Analyze historical data to predict storage growth.
• Business Requirements: Align with expected projects, applications, or expansions (e.g.,
database growth, application deployments, etc.).
• Regulatory Compliance: Plan for storage required to meet retention policies or data protection
regulations.
• Scalability Goals: Ensure the plan can accommodate unexpected surges or business pivots.
 Storage Capacity Planning and Expansion
3. Select the Right Storage Solutions
• Type of Storage:
• HDD: For cost-effective archival storage.
• SSD/NVMe: For high-performance needs.
• Object Storage: For unstructured data (e.g., AWS S3, Cloudflare R2).
• NAS/SAN: For shared storage in enterprise environments.

4. Plan for Expansion

• Scale-Out vs. Scale-Up:
• Scale-Up: Add capacity to existing systems (e.g., larger disks or RAID configurations).
• Scale-Out: Add more nodes or devices to the storage network.
• RAID Levels: Use RAID configurations (e.g., RAID5, RAID6) for redundancy and performance.
• Thin Provisioning: Allocate storage dynamically to prevent underutilization.
 Storage Capacity Planning and Expansion
5. Optimize Existing Storage
• Data Deduplication: Remove duplicate data to free up space.
• Compression: Compress files where applicable.
• Archiving: Move infrequently accessed data to archival storage.
• Storage Tiering: Use different storage types for hot, warm, and cold data.

6. Implement Monitoring and Alerts

• Use tools like Prometheus, Grafana, or vendor-specific solutions to:
• Track storage performance and utilization.
• Set alerts for critical thresholds (e.g., >80% utilization).
• Predict failures or performance degradation.
 Storage Capacity Planning and Expansion
7. Backup and Disaster Recovery
• Ensure adequate storage is allocated for backups.
• Use cloud or off-site solutions for disaster recovery.
• Test recovery processes regularly to ensure data integrity.

8. Budget and Vendor Management

• Cost Analysis: Evaluate the cost per GB/TB for different storage solutions.
• Vendor Relationships: Negotiate for better terms or bulk discounts.
• Support and Warranty: Ensure vendor SLAs meet your needs.
 Storage Capacity Planning and Expansion
9. Document and Review Regularly
• Capacity Plan: Maintain updated documentation of the storage infrastructure.
• Review Frequency: Perform quarterly or annual reviews to align with business growth.
 Zoning Configuration in Traditional SAN Architecture
View WWPN
#List WWPN On FC
•Switch1fcswitch1> switchshow
Index Port Address Media Speed State Proto
==================================================
0 0 010000 id N16 Online FC F-Port 10:00:00:10:9a:c3:2e:e7
1 1 010100 id N16 Online FC F-Port 10:00:00:10:9a:c3:2e:f3
2 2 010a00 id N16 Online FC F-Port 20:70:00:c0:ff:f1:c4:20
3 3 010b00 id 16G Online FC F-Port 24:70:00:c0:ff:f1:c4:20

#List WWPN On FC
• Switch2fcswitch1> switchshow
Index Port Address Media Speed State Proto
==================================================
0 0 010000 id N16 Online FC F-Port 10:00:00:10:9a:c3:2e:e8
1 1 010100 id N16 Online FC F-Port 10:00:00:10:9a:c3:2e:f4
2 2 010a00 id N16 Online FC F-Port 21:70:00:c0:ff:f1:c4:20
3 3 010b00 id 16G Online FC F-Port 25:70:00:c0:ff:f1:c4:20
 Zoning Configuration in Traditional SAN Architecture
Create Alias
•#On FC Switch 1
alicreate "Server1-1","10:00:00:10:9a:c3:2e:e7"
alicreate "Server2-1","10:00:00:10:9a:c3:2e:f3"
alicreate "STRG-Ctrl-A1","20:70:00:c0:ff:f1:c4:20"
alicreate "STRG-Ctrl-B1","24:70:00:c0:ff:f1:c4:20"

•#On FC Switch 2
alicreate "Server1-2","10:00:00:10:9a:c3:2e:e8"
alicreate "Server2-2","10:00:00:10:9a:c3:2e:f4"
alicreate "STRG-Ctrl-A2","21:70:00:c0:ff:f1:c4:20"
alicreate "STRG-Ctrl-B2","25:70:00:c0:ff:f1:c4:20"
 Zoning Configuration in Traditional SAN Architecture
Create Zoning
#On FC Switch 1
zonecreate "ZONE1","Server1-1;STRG-Ctrl-A1;STRG-Ctrl-B1"
zonecreate "ZONE2","Server2-1;STRG-Ctrl-A1;STRG-Ctrl-B1"

#On FC Switch 2
zonecreate "ZONE1","Server1-2;STRG-Ctrl-A2;STRG-Ctrl-B2"
zonecreate "ZONE2","Server2-2;STRG-Ctrl-A2;STRG-Ctrl-B2"
 Zoning Configuration in Traditional SAN Architecture
Create CFG
#On FC Switch 1
cfgcreate "cfg-fcswitch1","ZONE1;ZONE2"

#On FC Switch 2
cfgcreate "cfg-fcswitch2","ZONE1;ZONE2"

Enable CFG
#On FC Switch 1
cfgenable "cfg-fcswitch1"

#On FC Switch 2
cfgenable "cfg-fcswitch2"
 Thank You!
Dipen Khadka Santosh Parajuli
9864475529 9861585018
dipen@cyphertechnepal.com santosh@cyphertechnepal.com