Filestore)¶

1. Storage Type Comparison¶

1.1 Block vs File vs Object Storage¶

Type	Characteristics	Use Cases	AWS	GCP
Block	Low-level disk access	Databases, OS boot disk	EBS	Persistent Disk
File	Shared filesystem	Shared storage, CMS	EFS	Filestore
Object	HTTP-based, unlimited	Backup, media, logs	S3	Cloud Storage

1.2 Service Mapping¶

Feature	AWS	GCP
Block Storage	EBS (Elastic Block Store)	Persistent Disk (PD)
Shared File Storage	EFS (Elastic File System)	Filestore
Local SSD	Instance Store	Local SSD

2. Block Storage¶

2.1 AWS EBS (Elastic Block Store)¶

EBS Volume Types:

Type	Use Case	IOPS	Throughput	Cost
gp3	General purpose SSD	Up to 16,000	Up to 1,000 MB/s	Low
gp2	General purpose SSD (legacy)	Up to 16,000	Up to 250 MB/s	Medium
io2	Provisioned IOPS	Up to 64,000	Up to 1,000 MB/s	High
st1	Throughput-optimized HDD	Up to 500	Up to 500 MB/s	Low
sc1	Cold HDD	Up to 250	Up to 250 MB/s	Very low

# EBS 볼륨 생성
aws ec2 create-volume \
    --availability-zone ap-northeast-2a \
    --size 100 \
    --volume-type gp3 \
    --iops 3000 \
    --throughput 125 \
    --tag-specifications 'ResourceType=volume,Tags=[{Key=Name,Value=MyVolume}]'

# EC2에 볼륨 연결
aws ec2 attach-volume \
    --volume-id vol-1234567890abcdef0 \
    --instance-id i-1234567890abcdef0 \
    --device /dev/sdf

# 인스턴스 내에서 마운트
sudo mkfs -t xfs /dev/xvdf
sudo mkdir /data
sudo mount /dev/xvdf /data

# fstab에 추가 (영구 마운트)
echo '/dev/xvdf /data xfs defaults,nofail 0 2' | sudo tee -a /etc/fstab

2.2 GCP Persistent Disk¶

Persistent Disk Types:

Type	Use Case	IOPS (read)	Throughput (read)	Cost
pd-standard	HDD	Up to 7,500	Up to 180 MB/s	Low
pd-balanced	SSD (balanced)	Up to 80,000	Up to 1,200 MB/s	Medium
pd-ssd	SSD (high performance)	Up to 100,000	Up to 1,200 MB/s	High
pd-extreme	High IOPS SSD	Up to 120,000	Up to 2,400 MB/s	Very high

# Persistent Disk 생성
gcloud compute disks create my-disk \
    --zone=asia-northeast3-a \
    --size=100GB \
    --type=pd-ssd

# VM에 디스크 연결
gcloud compute instances attach-disk my-instance \
    --disk=my-disk \
    --zone=asia-northeast3-a

# 인스턴스 내에서 마운트
sudo mkfs.ext4 -m 0 -E lazy_itable_init=0,lazy_journal_init=0,discard /dev/sdb
sudo mkdir /data
sudo mount -o discard,defaults /dev/sdb /data

# fstab에 추가
echo UUID=$(sudo blkid -s UUID -o value /dev/sdb) /data ext4 discard,defaults,nofail 0 2 | sudo tee -a /etc/fstab

3. Snapshots¶

3.1 AWS EBS Snapshots¶

# 스냅샷 생성
aws ec2 create-snapshot \
    --volume-id vol-1234567890abcdef0 \
    --description "My snapshot" \
    --tag-specifications 'ResourceType=snapshot,Tags=[{Key=Name,Value=MySnapshot}]'

# 스냅샷 목록 조회
aws ec2 describe-snapshots \
    --owner-ids self \
    --query 'Snapshots[*].[SnapshotId,VolumeId,StartTime,State]'

# 스냅샷에서 볼륨 복원
aws ec2 create-volume \
    --availability-zone ap-northeast-2a \
    --snapshot-id snap-1234567890abcdef0 \
    --volume-type gp3

# 스냅샷 복사 (다른 리전)
aws ec2 copy-snapshot \
    --source-region ap-northeast-2 \
    --source-snapshot-id snap-1234567890abcdef0 \
    --destination-region us-east-1

# 스냅샷 삭제
aws ec2 delete-snapshot --snapshot-id snap-1234567890abcdef0

Automated Snapshots (Data Lifecycle Manager):

# DLM 정책 생성 (매일 스냅샷, 7일 보관)
aws dlm create-lifecycle-policy \
    --description "Daily snapshots" \
    --state ENABLED \
    --execution-role-arn arn:aws:iam::123456789012:role/AWSDataLifecycleManagerDefaultRole \
    --policy-details '{
        "ResourceTypes": ["VOLUME"],
        "TargetTags": [{"Key": "Backup", "Value": "true"}],
        "Schedules": [{
            "Name": "DailySnapshots",
            "CreateRule": {"Interval": 24, "IntervalUnit": "HOURS", "Times": ["03:00"]},
            "RetainRule": {"Count": 7}
        }]
    }'

3.2 GCP Snapshots¶

# 스냅샷 생성
gcloud compute snapshots create my-snapshot \
    --source-disk=my-disk \
    --source-disk-zone=asia-northeast3-a

# 스냅샷 목록 조회
gcloud compute snapshots list

# 스냅샷에서 디스크 복원
gcloud compute disks create restored-disk \
    --source-snapshot=my-snapshot \
    --zone=asia-northeast3-a

# 스냅샷 삭제
gcloud compute snapshots delete my-snapshot

Snapshot Schedule:

# 스케줄 정책 생성 (매일, 7일 보관)
gcloud compute resource-policies create snapshot-schedule daily-snapshot \
    --region=asia-northeast3 \
    --max-retention-days=7 \
    --start-time=04:00 \
    --daily-schedule

# 디스크에 스케줄 연결
gcloud compute disks add-resource-policies my-disk \
    --resource-policies=daily-snapshot \
    --zone=asia-northeast3-a

4. Volume Expansion¶

4.1 AWS EBS Volume Expansion¶

# 1. 볼륨 크기 수정 (온라인 가능)
aws ec2 modify-volume \
    --volume-id vol-1234567890abcdef0 \
    --size 200

# 2. 수정 상태 확인
aws ec2 describe-volumes-modifications \
    --volume-id vol-1234567890abcdef0

# 3. 인스턴스 내에서 파일시스템 확장
# XFS
sudo xfs_growfs -d /data

# ext4
sudo resize2fs /dev/xvdf

4.2 GCP Persistent Disk Expansion¶

# 1. 디스크 크기 확장 (온라인 가능)
gcloud compute disks resize my-disk \
    --size=200GB \
    --zone=asia-northeast3-a

# 2. 인스턴스 내에서 파일시스템 확장
# ext4
sudo resize2fs /dev/sdb

# XFS
sudo xfs_growfs /data

5. File Storage¶

5.1 AWS EFS (Elastic File System)¶

Features: - NFS v4.1 protocol - Auto-scaling (expand/shrink) - Multi-AZ support - Supports thousands of concurrent EC2 connections

# 1. EFS 파일 시스템 생성
aws efs create-file-system \
    --performance-mode generalPurpose \
    --throughput-mode bursting \
    --encrypted \
    --tags Key=Name,Value=my-efs

# 2. 마운트 타겟 생성 (각 서브넷)
aws efs create-mount-target \
    --file-system-id fs-12345678 \
    --subnet-id subnet-12345678 \
    --security-groups sg-12345678

# 3. EC2에서 마운트
sudo yum install -y amazon-efs-utils
sudo mkdir /efs
sudo mount -t efs fs-12345678:/ /efs

# 또는 NFS로 마운트
sudo mount -t nfs4 -o nfsvers=4.1 \
    fs-12345678.efs.ap-northeast-2.amazonaws.com:/ /efs

# fstab에 추가
echo 'fs-12345678:/ /efs efs defaults,_netdev 0 0' | sudo tee -a /etc/fstab

EFS Storage Classes: | Class | Use Case | Cost | |--------|------|------| | Standard | Frequent access | High | | Infrequent Access (IA) | Infrequent access | Low | | Archive | Long-term retention | Very low |

# 수명 주기 정책 설정 (30일 후 IA로 이동)
aws efs put-lifecycle-configuration \
    --file-system-id fs-12345678 \
    --lifecycle-policies '[{"TransitionToIA":"AFTER_30_DAYS"}]'

5.2 GCP Filestore¶

Features: - NFS v3 protocol - Pre-provisioned capacity - High-performance options available

Filestore Tiers: | Tier | Capacity | Performance | Use Case | |------|------|------|------| | Basic HDD | 1TB-63.9TB | 100 MB/s | File sharing | | Basic SSD | 2.5TB-63.9TB | 1,200 MB/s | High performance | | Zonal | 1TB-100TB | Up to 2,560 MB/s | High-performance workloads | | Enterprise | 1TB-10TB | Up to 1,200 MB/s | Mission-critical |

# 1. Filestore 인스턴스 생성
gcloud filestore instances create my-filestore \
    --zone=asia-northeast3-a \
    --tier=BASIC_SSD \
    --file-share=name=vol1,capacity=2.5TB \
    --network=name=default

# 2. Filestore 정보 조회
gcloud filestore instances describe my-filestore \
    --zone=asia-northeast3-a

# 3. VM에서 마운트
sudo apt-get install -y nfs-common
sudo mkdir /filestore
sudo mount 10.0.0.2:/vol1 /filestore

# fstab에 추가
echo '10.0.0.2:/vol1 /filestore nfs defaults,_netdev 0 0' | sudo tee -a /etc/fstab

6. Local SSD¶

6.1 AWS Instance Store¶

Instance Store is temporary storage physically attached to EC2 instances.

Features: - Data loss on instance stop/termination - Very high IOPS - No additional cost (included in instance price)

# Instance Store가 있는 인스턴스 유형 확인
aws ec2 describe-instance-types \
    --filters "Name=instance-storage-supported,Values=true" \
    --query 'InstanceTypes[*].[InstanceType,InstanceStorageInfo.TotalSizeInGB]'

# 예: i3.large, d2.xlarge 등

# 인스턴스 내에서 마운트
sudo mkfs.xfs /dev/nvme1n1
sudo mkdir /local-ssd
sudo mount /dev/nvme1n1 /local-ssd

6.2 GCP Local SSD¶

# Local SSD가 있는 인스턴스 생성
gcloud compute instances create my-instance \
    --zone=asia-northeast3-a \
    --machine-type=n2-standard-4 \
    --local-ssd=interface=NVME \
    --local-ssd=interface=NVME

# 인스턴스 내에서 마운트
sudo mkfs.ext4 /dev/nvme0n1
sudo mkdir /local-ssd
sudo mount /dev/nvme0n1 /local-ssd

Local SSD Features: - Added in 375GB units - Up to 24 units (9TB) - Data loss on instance stop - Live Migration not available (for some)

7. Performance Optimization¶

7.1 IOPS vs Throughput¶

IOPS (Input/Output Per Second):
- Number of read/write operations per second
- Important for small random I/O
- Databases, transaction processing

Throughput:
- Amount of data transferred per second (MB/s)
- Important for large sequential I/O
- Video streaming, big data

7.2 Optimization Tips¶

AWS EBS:

# gp3 IOPS/처리량 조정
aws ec2 modify-volume \
    --volume-id vol-xxx \
    --iops 10000 \
    --throughput 500

# EBS 최적화 인스턴스 사용
aws ec2 run-instances \
    --instance-type m5.large \
    --ebs-optimized \
    ...

GCP Persistent Disk:

# 더 큰 디스크 = 더 높은 성능
# pd-ssd 100GB: 최대 3,000 IOPS
# pd-ssd 500GB: 최대 15,000 IOPS
# pd-ssd 1TB: 최대 30,000 IOPS

# 성능을 위해 디스크 크기 증가
gcloud compute disks resize my-disk --size=500GB

8. Cost Comparison¶

8.1 Block Storage Cost (Seoul Region)¶

Type	AWS EBS	GCP PD
General Purpose SSD	$0.114/GB (gp3) \| $0.102/GB (pd-balanced)
High Performance SSD	$0.138/GB (io1) \| $0.180/GB (pd-ssd)
HDD	$0.054/GB (st1) \| $0.044/GB (pd-standard)

8.2 File Storage Cost¶

Service	Cost
AWS EFS Standard	~$0.33/GB/month
AWS EFS IA	~$0.025/GB/month
GCP Filestore Basic SSD	~$0.24/GB/month
GCP Filestore Basic HDD	~$0.12/GB/month

9. Hands-on: Setting Up Shared Storage¶

9.1 AWS EFS Multi-Instance Mount¶

# 1. 두 개의 서브넷에 마운트 타겟 생성
aws efs create-mount-target --file-system-id fs-xxx --subnet-id subnet-1 --security-groups sg-xxx
aws efs create-mount-target --file-system-id fs-xxx --subnet-id subnet-2 --security-groups sg-xxx

# 2. 보안 그룹에 NFS 규칙 추가
aws ec2 authorize-security-group-ingress \
    --group-id sg-xxx \
    --protocol tcp \
    --port 2049 \
    --source-group sg-instance

# 3. 각 인스턴스에서 마운트
# Instance 1
sudo mkdir /shared && sudo mount -t efs fs-xxx:/ /shared
echo "Hello from Instance 1" | sudo tee /shared/test.txt

# Instance 2
sudo mkdir /shared && sudo mount -t efs fs-xxx:/ /shared
cat /shared/test.txt  # "Hello from Instance 1" 출력

9.2 GCP Filestore Multi-Instance Mount¶

# 1. 방화벽 규칙 추가
gcloud compute firewall-rules create allow-nfs \
    --allow tcp:2049,tcp:111,udp:2049,udp:111 \
    --source-ranges 10.0.0.0/8

# 2. 각 인스턴스에서 마운트
# Instance 1
sudo mkdir /shared && sudo mount 10.0.0.2:/vol1 /shared
echo "Hello from Instance 1" | sudo tee /shared/test.txt

# Instance 2
sudo mkdir /shared && sudo mount 10.0.0.2:/vol1 /shared
cat /shared/test.txt  # "Hello from Instance 1" 출력

10. Next Steps¶

09_Virtual_Private_Cloud.md - VPC Networking
11_Managed_Relational_DB.md - Database Storage