February was super busy in term of AWS Release. So here a quick summary of what catch my eyes:

1) Testing:

Testing is super important in software development, so having so help for testing is really interesting

Step functions:

Step function is a service that helps you orchestrate you workflows really easily.

The big news here is that it now supports mocking for testing/debugging. That’s really interesting as it should help you testing your code change without impacting your production stage and without having to clone it for testing purpose.

Mock testing is not a perfect testing solution but having the option in unit test/early integ test is super helpful.

Fault Injection Simulator:

That’s a service I wasn’t aware of, and it seams super interesting as it can be really tricky to test a real life failure scenario to ensure you have valid mechanisms in place to handle the consequences are efficient.

S3/CloudWatch support:

It’s possible to review from S3 and/or CloudWatch the logs relative to the actions you created for you fault simulation.

Task Level Actions:

It’s possible to setup action that will shutdown specific tasks in container based application running in ECS.

Private Link:

Private link is a service that allows you to create a secured connection between VPCs AWS Services and on-premises networks.

We have 2 new services that supports private link:

• Elasticache
• MemoryDB for REDIS

Security:

Code Guru:

Code guru is service that will help you finding security issues in your code.

Detection of Log4J and other Log injections vulnerabilities:

Log4J was a really big vulnerability found in last December, it’s really being able to find those vulnerabilities is super important.

Detector Libraries and Examples:

This is really interesting as it will help you understand vulnerabilities better and mitigate the related risks. Example are currently available in Java and Python.

Secrets Manager:

Secrets manager is service that securely stores your secrets and allow to rotate them.

I now supports rotation windows that will help you define when your secrets need to be rotated to ensure the rotation happens when it will have less impact on your application.

WAF:

Rule Versioning:

Super important, it’s so easy to create issues in a firewall rule, so being able to audit and view the changes that happens will help for recovery if a faulty rule is created by simply rolling back.

Databases:

The big releases of the month are there:

Redshift:

Kinesis Data Stream Source:

Load live streams into Redshift can be quite complicated sometimes, before you needed to have some compute service to read the streams and output it into S3 before being able to load it into Redshift. New you can bypass the S3 step entirely and load the stream directly into Redshift in a matter of seconds.

Cross Region Data Sharing:

Quite nice too, you can share data between Redshift clusters in different region, that was pretty tedious previously as you needed to unload data into S3, transfer the data to another region then load the data. Now it’s possible to use Redshift data sharing service. I also notices that UNLOAD function in Redshift now support cross region unload as well?

Unload as JSon:

Sometime you want to represent you data in a JSon file as it’s easier to use with many applications, that’s why it’s now possible to UNLOAD data into JSON files.

Automatic Workload Management:

Managing your workload in Redshift can be complex, the new AutoWLM feature uses machine learning and it works with Concurrency Scaling.

RDS:

Postgres/MySQL:

Multi-AZ is now 2x faster for commit latency and that allow under 35seconds failover.

MySQL:

• Improve fail over for JDBC base application

MariaDB:

• Delayed Replication: You can implement a replication delay that allows you to switch master right after any issue (like dropping a table) to propagate further downstream.
• IAM Authentication support

Glue:

Glue now supports multiple file outputs, pretty usefull when you deal with huge amount of data that you want to split in smaller parts for processing.

S3

Batch replication:

This feature allows you to keep buckets in sync. You can select multiple object to ensure the data stays the closest to you customer;