Eastic File System¶
TLDR¶
High performance network file system which is pay per gb use and can attach to multiple services in multiple AZs.
Features¶
- multi-AZ
- high avalilable, scalable and expensive ( 3x gp2)
- pay per use
- uses SecurityGroup
- private service, via mount targets inside a docs/networks/VPC
- can be accessed from on-premises - VPN or DX
- General Purpose and Max I/O performance Modes
Uses cases¶
- Content Management, web serving, data sharing, wordpress
- NFSv4.1 protocol
- Only compatible with Linux based AMI
- encyption at rest with KMS
- posix file system
Performance¶
- 1000s of clients 10GBs throughput
- Can handle petabyte scale data
Performances Modes¶
- set at creation time
General purpose¶
- web server, cms , file serving
- latency sensitive
- default for 99.9% of uses
Max IO¶
- higher latency and throughput
- paralell
- big data, media processing
Throughput modes¶
- Bursting
- Provisioned
bursting¶
- 1 tb = 50mbs + burst of up to 100mbs
- like gp2 volumes inside a EBS
- throughput scale with the size of the file system
- generally default for wide variaty of cases
provisioned¶
- set throughput regardless of size
- more like io1 from EBS
- use when data created is small but performance needs to be high
Storage classes¶
Storage Tiers¶
- livecycle managemet for cost savings
- uses livecycle policies
Standard¶
- frequently access files
- default for almost all use cases
- data redundancy across multiples azs
Infrequent access¶
- cost to retrieve files but lower storage costs
- designed for infrequent store
Availablity¶
Standard¶
- multi AZ
One Zone¶
- one AZ
- good for testing env
- compatible with infrequent acess ( for up to 90% in savings)
- backup enabled by default