Amazon Announces Amazon DynamoDB Transactions, CloudWatch Logs Insights, And Firecracker

Amazon has made a number of new announcements at its AWS re:Invent 2018 conference to enable developers to get started with non-virtual environments, artificial intelligence, and gain deeper insights.

Amazon has made a number of new announcements at its AWS re:Invent 2018 conference to enable developers to get started with non-virtual environments, artificial intelligence, and gain deeper insights. 
Amazon Firecracker - Secure and Fast microVM for Serverless Computing
The company has launched Firecracker, an open-source solution that allows lightweight micro-virtual machines to be launched in non-virtual environments. Firecracker combines the security and workload isolation of traditional VMs with the resource efficiency of containers.
According to AWS, it uses multiple levels of isolation, resulting in minimal attack surface. It can also be used to power high-volume AWS services, such as AWS Lambda and AWS Fargate.
Amazon Re:Invent 
Source: Amazon 
The company also announced a new service - CloudWatch Logs Insights - to provide better insight into service log data. The service is able to work through massive logs in a short period of time and provide interactive queries and visualizations.
According to the company, it uses a sophisticated query language that features commands that can fetch specific event fields, filter based on conditions, calculate aggregate statistics, sort on the desired file, and limit the number of events that a query returns.
Support for Java in Amazon Kinesis Data Analytics
The company has also introduced support for Java in Amazon Kinesis Data Analytics, enabling developers to write their own Java code to create applications for processing streaming data.
"To use this new functionality, developers build applications using open source libraries which include built-in operators for common data processing functions that allow applications to organize, transform, aggregate, and analyze data at any scale. These libraries are both open source and you can run them anywhere" wrote Danilo Poccia of AWS.
There are two open-source libraries that can be used to build these applications: Apache Flink and AWS SDK for Java.
Poccia explained that now developers can use their chosen IDE and integrate with AWS services such as Amazon Kinesis Data Streams, Amazon S3, Amazon DynamoDB, and Amazon Kinesis Data Firehose.
Amazon DynamoDB transactions
Transactions have been added to DynamoDB, providing developers with atomicity, consistency, isolation, and durability (ACID).
Transactions in DynamoDB will enable use cases such as processing financial transactions, fulfilling and managing orders, building multiplayer game engines, and coordinating actions across distributed components and services.
Two new DynamoDB operations have been introduced for handling transactions.
TransactWriteItems, a batch operation that contains a write set, with one or more PutItem, UpdateItem, and DeleteItem operations. It can optionally check for prerequisite conditions that need to be satisfied before making updates.
TransactGetItems, a batch operation that contains a read set, with one or more GetItem operations. If this request is issued on an item that is part of an active write transaction, the read transaction is canceled.
"There is no additional cost to enable transactions for DynamoDB tables. You only pay for the reads or writes that are part of your transaction. " wrote the company.
Amazon SageMaker Neo improves machine learning processes
The company also announced Amazon SageMaker Neo, a new capability of Amazon SageMaker that enables machine learning models to train once and run anywhere in the cloud and at the edge with optimal performance.
"Without any manual intervention, Amazon SageMaker Neo optimizes models deployed on Amazon EC2 instances, Amazon SageMaker endpoints and devices managed by AWS Greengrass." explained Julien Simon of AWS.
Finally, Amazon announced a new solution that will make time series forecasting easier -Amazon Forecast - a fully managed deep learning service that can be used to generate predictions on time series data.