Preparation for certification

Those are the notes I took during the Cloud Guru AWS Certified Solutions Architect - Associate (SAA-C03). Note that the course content changes as the AWS changes. The notes are from March-May 2024.

This section is about AWS solutions that ease data migrations and/or offer hybrid storage.

AWS Snow Family

moving data to Cloud over Internet can have drawbacks (security, speed, reliability)
peta-scale, AWS ships physical hard drives to you, and you send tem back
types of services:
- Snowcone AWS DataSync
  - you also get a Compute to process and download your data first
  - smallest EC2, 8TB storage, 4GB memory, 2v2CPUs
  - direct IoT sensor integration
  - good for edge computing where power and space are constrained
- Snowball Edge
  - storage of 48-81TB
  - optionally small Compute, GPU flavors
  - e.g. on a boat where you don’t have good connection, or one time migration
  - for databases, can be used in conjunction with Schema Conversion Tool (SCT) to move data to S3, and DMS can move it from S3 to the AWS DB
    - still compatible with CDC (how?)
- Snowmobile
  - 100PB of data, exabyte-scale
in exam also data encrypted at rest and in transit could qualify for Snow Family
all offerings works both directions (to migrate to and from AWS)
- usually takes about a week

Storage Gateway

hybrid cloud storage service
can help with one time migration or to have the hybrid storage permanently
AWS provides VMs, transfer over HTTPS
offerings:
- File Gateway
  - network file share, NFS or SMB
  - data backup
  - put all data or still cache most recently used on-premise (Cached File Gateway)
    - helps when there is not enough storage
  - uses S3
- Volume Gateway
  - iSCSI mount
  - also cached or stored mode
  - uses EBS snapshots, you can restore the volume in AWS
- Tape Gateway
  - to move existing tape-based storage / backup to AWS
  - uses S3 / S3 GLacier
on exam, favor answers that mention complete migration

easily move files in and out of S3 / EFS, using SFTP, FTPS or FTP
the legacy client endpoint can still look like SFTP, FTPS or ~~FTP~~, but there is another service on AWS side
the DNS entry stays the same

a GUI to track the progress of SMS and DMS migrations
Server Migration Service (SMS)
- you have a VM in your datacenter, and want to move it to AWS
- you schedule when your data will be copied: S3 -> EBS snapshot -> AMI
Database Migration Service (DMS)
- in or out of AWS migration
- RDS and NoSQL
- one time migration (Full Load), or ongoing replication (Change Data Capture (CDC)), or both
  - CDC guarantees transactional integrity of the target DB
- you can use Schema Conversion Tool for translation to another DB engine
  - that tool supports OLAP and OLTP dbs, as well as data warehouses
  - on AWS side, you can import to any RDS, also RedShift
    - you can run the migrated DB even on EC2
- you have a DMS server which is running the migration software, you schedule tasks, you can create the schema yourself or leave it up to DMS
- the source and target are called endpoints
there are als third party solutions available
AWS Console: AWS Migration Hub -> Get started with discovery / Get started migrating
- discovery: import data, view discovered servers, group servers as applications
  - discovery agent - you install it on your VMs
  - discovery connector - you install it in your vMWare VCenter, it collects info about your VMs

helps planning of the migrating to AWS, by collecting your on-premise configuration and usage data
integrates with AWS Migration Hub, simplifies migration tracking
you can group discovered servers and view migration status by application
discovery types:
- agentless - via an agentless collector: OVA file within the VMWare vCenter
  - collects IP and MAC addresses, resource memory/CPU allocations, hostnames, ..
  - collects many utilisation data metrics, including disc I/O
- agent-based - install this agent on eah of the VMs and each of your servers
  - both Linux and Windows versions
  - collects more info: on top static config data, time-series performance info, network connections, OS processes

migrate your infrastructure and services to AWS with minimal changes
used to avoid cutover windows and disruptions
for physical, virtual or cloud servers
replicates source servers into AWS and automatically launches them when you are ready
Recovery Time Objective (RTO) - how quick to restore normal operations, depends on OS boot time, typically minutes
Recovery Point Objective (RPO) - how much data back in time can be lost, sub second range

integration service, for data exchange between AWS and SaaS app
- e.g. for migrating data from Salesforce to S3
you define how to ingest data and put them in an AWS service, or the other direction (bi-directional)
terms:
- Flow - transfers data between source and destination
- Data Mapping - how source data is stored in the destination
- Filters - criteria which data records are transferred to the destination
- Trigger - how the flow is started: on demand, on event or on schedule
first you establish a con nection
can use <100GB per flow of data transfer
use case examples:
- transferring Salesforce records to Redshift
- ingesting and analysing Slack conversations and storing them in S3
- migrating Zendesk support tickets to Snowflake
- transfer aggregated data on a scheduled basis to S3
- generally, easy and fast SaaS/third party data transfer from and to AWS, especially on schedule

Comments: