High Availability Flashcards
EC2
Scalability and Elasticity
Scalability
* Increase Instance Size as required, using reserved instances
Elasticity
* Increase the number of EC2 instances, based on autoscaling
DynamoDB
Scalability and Elasticity
Scalability
* Unlimited amount of storage
Elasticity
* Increase additional IOPS for additional spikes in traffic
* Decrease that IOPS after the spikeRDS
Scalability and Elasticity
Scalability
* Increase Instance Size as required.
* from small to medium
* There is a small downtime, do it during maintenance window
Elasticity
* Can’t scale RDS based on demand
Aurora
Scalability and Elasticity
Scalability
* Modify instace type, small to large
Elasticity
* Aurora Serverless
Multi-AZ
- Used for Failover
- Having a synchronized copy of your Primary DB in another AZ
- It gets the endpoint and points to your other AZ
Multi-AZ for SQL Server
- Engine uses synchronous logical replication
* It uses SQL Server-native Mirroring technology
Multi-AZ for MySQL, Oracle, and PostgreSQL
- Engine uses synchronous physical replication
You are doing a backup and restore, it’s affecting your production database, how can you avoid that?
- By enabling Multi-AZ:
* Backups and restores are taken from secondary DB which avoids I/O suspension to the primary
Read Replicas
It’s a read-only database.
When would you use Read Replicas?
- Scaling beyond the compute or I/O capacity of a single DB instance for read heavy workloads.
- Serving read traffic while the source DB Instance is unavailable.
- Business reporting or data warehousing scenarios:
- Run your BI queries against Read Replicas, rather than primary DB Instance
- Or use redshift to run these queries
What happens when you create a new Read Replica?
- AWS takes a snapshot of your db
- Multi-AZ enabled: NO downtime
- MUlti-AZ not enabled: Downtime
- RDS manages endpoints and all
True or False
You can have read replicas in different Regions for all engines
True
True or False
Read Replicas cannot be multi-AZ
False - Read Replicas can be multi-AZ
What is a key metric to look for using Read Replicas?
REPLICA LAG
The higher, the worst
When is a good use case for promoting read replicas into its own DB instance?
- When you want to test, develop, run BI queries on your DB
* It’s like cloning your DB instance.
How do you simulate a failover?
Reboot the DB so it does the failover.
What are the steps to encrypt RDS Instances?
- Take a snapshot of existing RDS instance
- Copy the snapshot to the same/different region.
- Encrypt the copy during the copy process
- Restore the snap
Sharing Encrypted RDS Snapshots between accounts
- Create a CUSTOM KMS Encryption key.
- Create an RDS snapshot using the custom key
- Share the CUSTOM AWS KMS encryption key that was used to encrypt the snapchot
- Use the Console, CLI, or Amazon RDS API to share the encrypted snapshot with the other account.
What happens when Aurora is at 100% CPU Utilization?
Writes causing issue.
Scale Up → Increase instance size
What happens when Aurora is at 100% CPU Utilization?
Reads causing issue.
Scale Out → Add more read replicas
aws rds describe-db-instances –region
What does this command do?
See what version you are using, and all the description for the instance that you are using.
Which Services have Maintenance Windows?
RERDND
- RDS
- Elasticache
- Redshift
- DynamoDB DAX
- Neptune
- Amazon DocumentDB
Services without maintenance windows:
- EC2
- Lambda
- QLDB
Which engine of ElastiCache support Multi-AZ?
Redis - Master/Slave replication
- Memcached does not support Multi-AZ
_____ ______ _______ is the number of times that your edge locations are able to serve content back to your user without having to go back to the origin. The more requests from edge locations, the better.
Cache Hit Rations (CloudFront)
Where can I visualize the Cache Hit Ratio?
On the CloudFront Console.
To increase your cache hit ratio, you can configure your origin to add a Cache-Control: max-age directive to your objects, and specify the longest practical value for a max-age. So, the shorter the cache duration, the more frequently CloudFront forwards another request to your origin to determine whether the object has changed, and, if so, to get the latest version.
Specifying how long CloudFront Caches your objects. (TTL)
What is important to make sure of when Caching based on Query String Parameters?
Make sure that you do consistent naming convention across your application for variables using query string parameters.
http: //www.example.com?id=a1
http: //www.example.com?id=A1
http: //www.example.com?Id=a1
http: //www.example.com?ID=a1
These would generate 4 different requests to the origin, because the parameter is case sensitive.
Caching based on Cookie Values, how should you configure?
- Create separate cache behaviors for static and dynamic content
- Configure CloudFront to forward cookies to your origin only for dynamic content.
6 ways to improve CloudFront Cache Hit Ratio:
- Specifying how long CloudFront Caches your objects
- Caching based on Query String Parameters
- Caching based on Cookie Values
- Caching based on Request Headers
- Remove Accept-Enconding Headers when compression is not needed
- Serving media content by using HTTP
Caching based on Request Headers, how should you configure?
- Configure CloudFront to forward and cache based only on specified headers instead of forwarding and caching based on all headers.
- Avoid caching based on request headers that have large numbers of unique values.
How can you serve media content by using HTTP?
You can use CloudFront to deliver video on demand or live streaming video using any HTTP origin.
Use CloudFront together with AWS Media Services.
What does list of things may show you?
- Associated key pair does not exist
- Security Group does not exist
- Autoscaling config is not working correctly
- Austoscaling group not found
- Instance type specified is not supported in the AZ
- AZ is no longes supported
- Invalid EBS device mapping
- Autoscaling service is not enabled on your account
- Attempting to attach an EBS block device to an instance-store AMI
The reason for my instance not launching into an autoscaling group.
Aurora Cross Region Replicas:
- Creating a new cross region replica will also create a new Aurora cluster in the target region.
- If replications is disrupted, you will have to set up again.
- It is recommended that you select Multi-AZ deployment to ensure high availability for the target cluster.
Aurora Encryption
- Turned on by default
* All read replicas will be encrypted
Aurora Failover
- It is defined by tiers
- The lower the Tier, higher the priority.
- Tier 0 → highest priority
Amazon ElastiCache can fulfill a number of roles. What are thee things which can be implemented using Elasticache for Redis.
- Sorted Sets
- In Memory Data Store
- Pub/Sub
How can you implement redundancy for your RDS DB instances?
Redundancy = High Availability
- Using Multi-AZ Deployments
