Amazon EFS
Amazon EFS Overview
- Definition: Amazon Elastic File System (EFS) is a fully managed, scalable file storage service that provides a simple, serverless NFS (Network File System) for use with AWS Cloud services and on-premises resources, designed for shared access across multiple EC2 instances or containers.
- Key Features:
- Supports NFSv4.0/4.1, enabling shared access for thousands of connections.
- Offers Standard and Infrequent Access (IA) storage classes for cost optimization.
- Provides elastic capacity, scaling automatically without provisioning.
- Integrates with AWS Backup, VPC, and IAM for robust management.
- Use Cases: Content management, web serving, home directories, media processing, container storage, big data analytics.
1. EFS Core Concepts
Components
- File System:
- Logical container for files and directories, accessible via NFS.
- Regional resource, spans multiple AZs for high availability.
- Explanation: E.g., file system for shared web content.
- Mount Target:
- Network endpoint (IP address) in a VPC subnet for accessing the file system.
- One mount target per AZ for Multi-AZ access.
- Explanation: E.g., mount target in us-east-1a for EC2 access.
- Access Point:
- Application-specific entry point with enforced root directory and permissions.
- Simplifies container and Lambda access.
- Explanation: E.g., access point for /app1 with read-only access.
- Storage Classes:
- Standard: Low-latency, high-throughput access.
- Infrequent Access (IA): Lower-cost, for rarely accessed data.
- Explanation: E.g., Standard for active files, IA for archives.
- Lifecycle Policies:
- Automatically move files between Standard and IA based on access patterns.
- Explanation: E.g., move files to IA after 30 days without access.
Performance Modes
- General Purpose (default):
- Low latency, suitable for most workloads (e.g., web servers, CMS).
- Explanation: E.g., use for latency-sensitive apps.
- Max I/O:
- Higher throughput, higher latency, for parallel processing (e.g., big data).
- Explanation: E.g., use for Hadoop analytics.
Throughput Modes
- Bursting:
- Scales throughput with file system size (e.g., 100 MiB/s baseline per TB).
- Uses burst credits for spikes.
- Explanation: E.g., bursts to 300 MiB/s for small file systems.
- Provisioned:
- Fixed throughput (up to 20 GiB/s), independent of size.
- Explanation: E.g., provision 1 GiB/s for media processing.
- Elastic (new):
- Automatically scales throughput with demand, no provisioning needed.
- Explanation: E.g., auto-scale for unpredictable workloads.
Recent Updates (2024–2025):
- Elastic Throughput: Auto-scaling throughput for unpredictable workloads (October 2024).
- Lower-Cost IA Storage: Reduced pricing for Standard-IA and One Zone-IA (March 2024).
- AWS Backup Enhancements: PITR, cross-region replication for EFS (April 2024).
- Security Enhancements: Default encryption, fine-grained IAM policies (February 2024).
Key Notes:
- Exam Relevance: Understand file systems, mount targets, storage classes, and throughput modes.
- Mastery Tip: Compare EFS vs. EBS vs. FSx for specific workloads.
2. EFS Performance Features
EFS supports high-performing shared storage.
Scalability
- Purpose: Handle dynamic workloads.
- Features:
- Elastic capacity (1 GiB to petabytes) without provisioning.
- Supports thousands of concurrent connections.
- Explanation: E.g., scale from 10 GB to 1 PB for media storage.
High Throughput
- Bursting: Up to 100 MiB/s baseline, bursts to 300 MiB/s per TB.
- Provisioned: Up to 20 GiB/s (e.g., 1 GiB/s for 100 GB file system).
- Elastic: Auto-scales to meet demand (no burst credits needed).
- Explanation: E.g., Elastic mode for spiky web traffic.
Low Latency
- General Purpose: Sub-millisecond latency for most workloads.
- Explanation: E.g., fast access for CMS files.
Access Points
- Purpose: Simplify app access.
- Features: Enforce directory paths, permissions for containers/Lambda.
- Explanation: E.g., Lambda reads /data via access point.
Key Notes:
- Performance: Elastic throughput + access points = flexible, fast storage.
- Exam Tip: Know when to use Elastic vs. Provisioned throughput.
3. EFS Resilience Features
Resilience ensures data availability and recovery.
Multi-AZ Architecture
- Purpose: High availability.
- Features:
- Standard storage class spans multiple AZs in a Region.
- Mount targets in each AZ for failover.
- Explanation: E.g., EC2 in us-east-1a accesses EFS if us-east-1b fails.
One Zone Storage Class
- Purpose: Cost-effective resilience.
- Features:
- Data stored in a single AZ, lower cost (~$0.043/GB-month vs. $0.30).
- Suitable for non-critical workloads.
- Explanation: E.g., use One Zone-IA for dev backups.
Backups with AWS Backup
- Purpose: Disaster recovery.
- Features:
- Policy-driven backups, PITR, cross-region replication.
- Daily backups with retention (e.g., 35 days).
- Explanation: E.g., restore EFS to 2 hours ago after accidental deletion.
Data Durability
- Purpose: Prevent data loss.
- Features: 99.999999999% (11 9s) durability for Standard class.
- Explanation: E.g., reliable storage for critical files.
Key Notes:
- Resilience: Multi-AZ + AWS Backup = robust file storage.
- Exam Tip: Design HA with Standard class and PITR backups.
4. EFS Security Features
Security aligns with SAA-C03’s secure architecture focus.
Encryption
- At Rest:
- Default encryption with KMS (AWS-managed or customer-managed keys).
- Explanation: E.g., encrypt CMS files with custom KMS key.
- In Transit:
- TLS for NFS connections (mandatory since February 2024).
- Explanation: E.g., secure file transfers between EC2 and EFS.
Access Control
- IAM:
- Controls file system access (e.g., elasticfilesystem:ClientMount).
- Fine-grained policies for access points (e.g., read-only for Lambda).
- Example: {"Effect": "Allow", "Action": "elasticfilesystem:ClientRead", "Resource": "arn:aws:elasticfilesystem:us-east-1:123456789012:access-point/fsap-123"}.
- POSIX Permissions:
- Standard Linux permissions (user, group, mode).
- Explanation: E.g., set 644 for read-only web files.
- VPC Security:
- Mount targets in private subnets, secured by security groups.
- Explanation: E.g., allow port 2049 (NFS) from EC2 security group.
Compliance
- Certifications: HIPAA, PCI, SOC, ISO, GDPR.
- Explanation: E.g., deploy healthcare app with encrypted EFS.
Key Notes:
- Security: KMS + IAM + TLS = compliant file storage.
- Exam Tip: Practice IAM policies and security group configs for EFS.
5. EFS Cost Optimization
Cost efficiency is a key exam domain.
Pricing
- Storage:
- Standard: ~$0.30/GB-month.
- Standard-IA: ~$0.013/GB-month.
- One Zone: ~$0.043/GB-month.
- One Zone-IA: ~$0.008/GB-month.
- Requests:
- ~$0.06/1000 read/write requests (Standard).
- ~$0.01/1000 for IA.
- Throughput:
- Provisioned: ~$6.00/MiB/s-month.
- Elastic: Pay-as-you-go, scales with usage.
- Data Transfer: Free within Region; charged for cross-region or on-premises.
- Free Tier: None.
- Example: 100 GB Standard, 50 GB Standard-IA, Elastic throughput ~$30.65/month.
Cost Strategies
- Infrequent Access:
- Use lifecycle policies to move files to IA after 7–30 days.
- Explanation: E.g., move logs to Standard-IA after 14 days.
- One Zone:
- Use for non-critical data to save ~85% vs. Standard.
- Explanation: E.g., One Zone-IA for backups.
- Elastic Throughput:
- Avoid over-provisioning with auto-scaling throughput.
- Explanation: E.g., Elastic for unpredictable CMS traffic.
- Tagging:
- Use cost allocation tags to track file system costs.
- Explanation: E.g., tag file system with “Project:Web”.
- Clean Up:
- Delete unused file systems to avoid charges.
Key Notes:
- Cost Savings: IA + One Zone + Elastic = low costs.
- Exam Tip: Calculate costs for Standard vs. IA and lifecycle policies.
6. EFS Advanced Features
Lifecycle Policies
- Purpose: Automate cost savings.
- Features:
- Transition files to IA or One Zone-IA based on last access.
- Options: 7, 14, 30, 60, 90, 180, 270, 365 days.
- Explanation: E.g., move media files to IA after 30 days.
Access Points
- Purpose: Secure, simplified access.
- Features:
- Enforce root directory, UID/GID, and permissions.
- Ideal for ECS, EKS, Lambda.
- Explanation: E.g., ECS task accesses /app via access point.
Cross-Region Replication
- Purpose: Disaster recovery.
- Features:
- Replicate file system to another Region via AWS Backup.
- Explanation: E.g., replicate EFS to us-west-2 for DR.
On-Premises Access
- Purpose: Hybrid cloud.
- Features:
- Access EFS via Direct Connect or VPN.
- Requires mount target in VPC.
- Explanation: E.g., on-premises server mounts EFS for shared storage.
Key Notes:
- Flexibility: Access points + replication = versatile file storage.
- Exam Tip: Know access points for containers and lifecycle policies for cost.
7. EFS Use Cases
Understand practical applications.
Content Management
- Setup: EFS + EC2 + General Purpose.
- Features: Shared access, low latency.
- Explanation: E.g., WordPress media storage.
Container Storage
- Setup: EFS + EKS + access points.
- Features: Persistent volumes, secure access.
- Explanation: E.g., Kubernetes pods share /data.
Big Data Analytics
- Setup: EFS + Max I/O + Provisioned.
- Features: High throughput, parallel access.
- Explanation: E.g., Spark cluster processes large datasets.
Home Directories
- Setup: EFS + One Zone + IA.
- Features: Cost-effective, scalable.
- Explanation: E.g., user home directories for enterprise.
8. EFS vs. Other Storage Services
Feature | EFS | EBS | S3 |
---|---|---|---|
Type | File Storage | Block Storage | Object Storage |
Workload | Shared file systems | Databases, boot volumes | Backups, archives |
Performance | Scalable, shared | High IOPS, low latency | High durability |
Cost | $0.008–$0.30/GB-month | $0.015–$0.125/GB-month | $0.023/GB-month |
Use Case | CMS, containers | EC2 storage | Data lakes |
Explanation:
- EFS: Shared, scalable file storage for multiple instances.
- EBS: High-performance block storage for single EC2.
- S3: Durable object storage for unstructured data.
Detailed Explanations for Mastery
- Elastic Throughput:
- Example: Auto-scale throughput for a CMS during traffic spikes.
- Why It Matters: Cost/performance optimization—new for 2024.
- Access Points:
- Example: Restrict EKS pod to /app with read-only access.
- Why It Matters: Security for containers—exam favorite.
- Lifecycle Policies:
- Example: Move logs to IA after 14 days, saving ~95% on storage.
- Why It Matters: Cost optimization—common scenario.
Quick Reference Table
Feature | Purpose | Key Detail | Exam Relevance |
---|---|---|---|
File System | Shared storage | NFSv4, Multi-AZ | Core Concept |
Mount Target | Network access | One per AZ, VPC subnet | Core Concept |
Elastic Throughput | Auto-scale performance | No provisioning, pay-as-you-go | Performance |
Multi-AZ | High availability | Standard class spans AZs | Resilience |
Encryption/IAM | Security | KMS, TLS, fine-grained policies | Security |
IA/One Zone | Cost savings | $0.008–$0.043/GB-month | Cost |
Access Points | Secure app access | Enforce directory/permissions | Flexibility |
AWS Backup | Disaster recovery | PITR, cross-region replication | Resilience |