Can I stop the replicas of a DynamoDB global table from synchronising? - amazon-web-services

I am having DynamoDB table in a specific region but the data it contains support application instances in multiple regions. I want to create a DDB per region setup without downtime.
In the end I want to have multiple instances running, each one in it's own region with it's own regional database table, but I also want the two tables to be in sync while the migration is rolling out.
I know that I can use DynamoDB streams with lambda to keep the two tables in sync for as long as I need, but I wonder if there's an easier way.
The idea is to add the extra region to the existing table, making it a global table. This will allow each local instance to use it's local database while also keeping the data in sync among regions.
But I don't want to maintain a global table for ever since after the migration is completed there's no reason to keep the replicas in sync.
So, is it possible to stop the replicas of a global table from syncing?
Is is possible to split a global table to it's local parts?
I couldn't find anything in the docs, but maybe I missed something.

Related

DynamoDB single-active cross region replication

I'm trying to understand DynamoDB replication & failover strategies but do not find any articles on the web which clarifies them. I understand cross-region replication can be achieved by DynamoDB with Global Tables but I also understand this is a multi-active table setup, meaning there are multiple active tables and multiple replica table. Is there a setup with single-active table and multiple replicas? I briefly read about this in this article but do not find any mentions anywhere else including AWS documentation.
I'm also trying to understand failover strategies for both cases - Is there a DynamoDB Java Client which can failover across AZs in case of issues in one AZ for both reads & writes?
DynamoDB Global Tables are always active-active but you can treat it as active-passive if you prefer. Many people do. That's useful if you want to use features like condition expressions, transactions, or do any non-idempotent wheres where you could have the same item being written around the same time in both regions with the second write happening before the first replicates, because this would cause the first write to be effectively lost.
To do this you just route your write traffic to one region, and to failover you decide when it's time to write to another. The failover region is always happy to be an active region if you'll let it.
As for AZs, DynamoDB is a regional service meaning it crosses at least 3 AZs always and would keep operating fine even if a full AZ were to be down. You don't have to worry about that.
Is there a setup with single-active table and multiple replicas
Unfortunately there is no such single active and multiple replica setup for cross region in dynamodb using global tables, so failover strategy will be for multiple active tables and multiple replica tables! - Source - docs
Fro failover strategies
According to docs
If a single AWS Region becomes isolated or degraded, your application can redirect to a different Region and perform reads and writes against a different replica table.
This means this is seamingless smooth process which happends by default ofcourse you can add custom logic when to redirect

DynamoDB Accelerator (DAX) cluster design: Per table cluster OR a single unified cluster?

We're having hundreds of DynamoDB tables.
For the performance optimization, we're going to use DynamoDB Accelerator (DAX).
While exploring DAX, I came across two approaches.
A unified cache cluster, that can be used for all DynamoDB tables
Separate cluster for each DynamoDB table
At a first glance, it seems #2 may be better because of isolation of individual clusters, as none of DynamoDB table's cluster will affect other table's cluster. However, manageability may be bit complex!
Is that correct OR am I missing anything? Which approach would be better and why?
Finally, We have used synthesis of both the approaches to get the merit of both approaches. Sharing it if it can help others!
To elaborate, multiple clusters are being created, and each cluster has been used for different set of DynamoDB tables.
Last note, remember that only one node from cluster handles write operation to DynamoDB and rest of nodes are just read replicas. Hence, while determining set of tables for a cluster, it should be considered.

DynamoDB Global Table Backup & Restore

I have a global DynamoDB table that is currently replicated across 3 regions (eu-west-1, eu-west-2, eu-central-1).
As part of a PoC piece of work I am looking to use AWS Backup to schedule automated backups, I was wondering what the best practice for this was?
Is it acceptable to take backups of a single region, i.e only schedule the backups for the table in eu-west-1? Then when it comes to recovering the table, I can go through the process of first restoring to a non-global table, then adding replica's.
Or is it better practice to ensure all region's tables are backed up at the same time?
I would suggest that you backup from a single region (perhaps if you have a primary region for writes use this.
If you restore the DynamoDB table, it needs to create a new DynamoDB table resource. Once this is restored you would then add your global tables which would replicate the data currently stored in the restored DynamoDB table.
By having multiple backups, you would need to have a strategy to preprocess for any differences between all regions your table exists.

Migrating Dynamodb Data between regions

I've a requirement to copy and move (depend on the case) items between AWS Dyanmodb regions.
For example, let's say, there is a USER table in ap-south-1 region and I want to move some items of it to the us-east-1 region.
So is there any way to migrate the data from one region to another considering large data set in mind?
I've read this solution [https://tvernon.tech/blog/move-dynamodb-data-between-regions]. But I am not sure how feasible it would be with large dataset.
See, I am not talking about Global tables here, which gives multi-region replication as a feature.

Amazon DynamoDB - geographically distributed?

I am new to AWS. Sorry if my question is basic, got stuck with this term.
AWS Global Infrastructure says "18 geographic Regions" -> Geographic term is used along with Regions, that makes sense.
DynamoDB FAQs 3rd questions says, "Amazon DynamoDB stores three geographically distributed replicas of each table to enable high availability and data durability."
Here(three geographically) is it referring to Region or Availability Zones ? Bit confused. If it is Region, does it mean my data is going out of my country(if my country has only 1 Region).
Please suggest.
Geographically isolated in this documentation refers to Availability Zones and not Regions. As per AWS documentation when you create a table in one region, it's replicated in others zones to ensure the high availability. If you do some activity in the table it's updated in the replicas. The AZ's are interconnected with low latency networks.
The data is stored on SSD disks and automatically replicated across
multiple Availability Zones in an AWS region, which brings the high
availability and your data is durable.
If you create a table in one region, the same table can be created in other regions also with same name.
If you want your table to be replicated in other regions you must enable the Cross-Region replication. For more details Refer
DynamoDB
All Things about DynamoDB
Almost every AWS service revolves around two things in availability: Multi AZ (multiple data centers in a single region) and Cross-Region (different geographic locations across globe) and so does the DynamoDB. By default AWS DynamoDB is a multi-AZ enabled service which means that your data is by default replicated across 3 data centers (minimum of 2 AZs) but for cross-region, you need to enable DynamoDB global tables (DynamoDB Streams).
Multi-Region Replication with DynamoDB
DynamoDB global tables are geographically distributed. They provide a fully managed solution for deploying a multiregion, multi-active database. Like with every other geographically distributed database, GlobalTables comes with ReplicationLatency.
An important thing to note here is, DynamoDB does not offer cross-region strong consistency (this is in contrast with CosmosDB, a similar offering from Azure)
From AWS documentation:
An application can read and write data to any replica table. If your
application only uses eventually consistent reads and only issues
reads against one AWS Region, it will work without any modification.
However, if your application requires strongly consistent reads, it
must perform all of its strongly consistent reads and writes in the
same Region. DynamoDB does not support strongly consistent reads
across Regions. Therefore, if you write to one Region and read from
another Region, the read response might include stale data that
doesn't reflect the results of recently completed writes in the other
Region.
Also, global tables are not to be confused with global indexes. Global indexes get their name because they are used in fetching data across multiple DynamoDB partitions.
"Amazon DynamoDB stores three geographically distributed replicas of each table to enable high availability and data durability."
This is specifically referring to multi AZ structure of dynamo, this helps in achieving high availability of your table. eg. if one of availability zone is down you still will be able to access you table.
To answer "my data is going out of my country(if my country has only
1 Region)."
For multi region its not by default ON you need to use global tables and specify regions in which you want to replicate that means your data/table wont go in any other region till you specifically want it to be.
For more on global tables refer
https://aws.amazon.com/dynamodb/global-tables/