Introducing new capabilities to AWS CloudTrail Lake to boost your cloud visibility and investigations

As we speak, I’m excited to announce new updates to AWS CloudTrail Lake, which is a managed knowledge lake you need to use to mixture, immutably retailer, and question occasions recorded by AWS CloudTrail for auditing, safety investigation, and operational troubleshooting.

The brand new updates in CloudTrail Lake are:

• Enhanced filtering choices for CloudTrail occasions
• Cross-account sharing of occasion knowledge shops
• Common availability of the generative AI–powered pure language question era
• AI-powered question outcomes summarization functionality in preview
• Complete dashboard capabilities, together with a high-level overview dashboard with AI-powered insights (AI-powered insights is in preview), a collection of 14 pre-built dashboards for varied use instances, and the flexibility to create customized dashboards with scheduled refreshes

Let’s look into the brand new options one after the other.

Enhanced filtering choices for CloudTrail occasions ingested into occasion knowledge shops
Enhanced occasion filtering capabilities offer you larger management over which CloudTrail occasions are ingested into your occasion knowledge shops. These enhanced filtering choices present tighter management over your AWS exercise knowledge, enhancing the effectivity and precision of safety, compliance, and operational investigations. Moreover, the brand new filtering choices aid you scale back your evaluation workflow prices by ingesting solely probably the most related occasion knowledge into your CloudTrail Lake occasion knowledge shops.

You may filter each administration and knowledge occasions based mostly on attributes akin to eventSource, eventType, eventName, userIdentity.arn, and sessionCredentialFromConsole.

I’m going to the AWS CloudTrail console and select Occasion knowledge shops below Lake within the navigation pane. I select Create occasion knowledge retailer. In step one, I enter a reputation within the Occasion knowledge retailer title area. For this demo, I depart different fields as default. You may select the pricing and retention choices that fit your wants. Within the subsequent step, I select Managements occasions and Information occasions below CloudTrail occasions. You may embrace all of the choices you want below CloudTrail occasions. You even have the choice to decide on ingestion choices. I select Ingest occasions to begin ingesting when it’s created. There could also be eventualities, while you need to deselect the Ingest occasions choice to cease an occasion knowledge retailer from ingesting occasions. For instance, chances are you’ll be copying path occasions to the occasion knowledge retailer and don’t need the occasion knowledge retailer to gather any future occasions. You may also select to allow ingestion for all accounts in your group or embrace solely the present area in your occasion knowledge retailer.

The next instance reveals an out of the field template for filtering, which excludes any administration occasions which are initiated by an AWS Service. I select Superior occasion assortment below the Administration occasions. I select Exclude AWS service-initiated occasions from the Log selector template dropdown. You may also broaden the JSON view to see how the filters really apply.

Underneath the Information occasions, the next instance creates a filter to incorporate DynamoDB knowledge occasions initiated by a sure consumer, serving to me to log occasions based mostly on an IAM principal. I select DynamoDB as Useful resource sort. I select Customized as Log selector template. Underneath the Superior occasion selector, I select userIdentity.arn as Discipline and equals as Operator. I enter the consumer’s ARN as Worth. I select Subsequent and select Create occasion knowledge retailer within the closing step.

Now, I’ve my occasion knowledge retailer that offers me granular management over the ingested CloudTrail knowledge.

This expanded set of filtering choices lets you be extra selective in capturing solely probably the most related occasions to your safety, compliance, and operational wants.

Cross-account sharing of occasion knowledge shops
You should use the cross-account sharing characteristic of occasion knowledge shops to boost collaborative evaluation inside organizations. It permits safe sharing of occasion knowledge shops with chosen AWS principals via Useful resource-Based mostly Insurance policies (RBP). This performance permits approved entities to question shared occasion knowledge shops throughout the identical AWS Area the place they have been created. 

To make use of this characteristic, I’m going to the AWS CloudTrail console and select Occasion knowledge shops below Lake within the navigation pane. I select an occasion knowledge retailer from the listing and navigate to its particulars web page. I select Edit within the Useful resource coverage part. The next instance coverage features a assertion that enables root customers in accounts 111111111111, 222222222222, and 333333333333 to run queries and get question outcomes on the occasion knowledge retailer owned by account ID 999999999999. I select Save modifications to avoid wasting the coverage.

Generative AI–powered pure language question era in CloudTrail Lake is now typically accessible
In June, we introduced this characteristic for CloudTrail Lake in preview. With this launch, you may generate SQL queries utilizing pure language questions to simply discover and analyze AWS exercise logs (solely administration, knowledge, and community exercise occasions) without having technical SQL experience. The characteristic makes use of generative AI to transform pure language questions into ready-to-use SQL queries you may run immediately within the CloudTrail Lake console. This simplifies the method of exploring occasion knowledge shops and retrieving insights akin to error counts, high providers used, and the causes of errors. This characteristic can be accessible via the AWS Command Line Interface (AWS CLI), offering further flexibility for customers preferring command-line operations. The preview weblog put up offers step-by-step directions on get began with the pure language question era characteristic in CloudTrail Lake.

CloudTrail Lake generative AI–powered question outcomes summarization functionality in preview
Constructing on the potential of pure language question era, we’re introducing a brand new AI-powered question outcomes summarization characteristic in preview to additional simplify the method of analyzing AWS account exercise. With this characteristic, you may simply extract invaluable insights out of your AWS exercise logs (solely administration, knowledge, and community exercise occasions) by mechanically summarizing the important thing factors out of your question ends in pure language, decreasing the effort and time required to know the knowledge.

To do that characteristic, I’m going to the AWS CloudTrail console and select Question below Lake within the navigation pane. I select an occasion knowledge retailer for my CloudTrail Lake question from the dropdown listing in Occasion knowledge retailer. You should use summarization no matter whether or not the question was written manually or generated by generative AI. For this instance, I’ll use the pure language question era functionality. Within the Question generator, I enter the next immediate within the Immediate area utilizing pure language:

What number of errors have been logged through the previous month for every service and what was the reason for every error?

Then, I select Generate question. The next SQL question is mechanically generated:

SELECT eventsource,
    errorcode,
    errormessage,
    depend(*) as errorcount
FROM a0******
WHERE eventtime >= '2024-10-14 00:00:00'
    AND eventtime <= '2024-11-14 23:59:59'
    AND (
        errorcode IS NOT NULL
        OR errormessage IS NOT NULL
    )
GROUP BY 1,
    2,
    3
ORDER BY 4 DESC;

I select Run to get the outcomes. To make use of the summarization functionality, I select Summarize outcomes within the Question outcomes tab. CloudTrail mechanically analyzes the question outcomes and offers a pure language abstract of the important thing insights. It’s necessary to notice that there’s a month-to-month quota of three MB for question outcomes that may be summarized.

This new summarization functionality can prevent effort and time in understanding your AWS exercise knowledge by mechanically producing significant summaries of the important thing findings.

Complete dashboard capabilities
Lastly, let me let you know concerning the new dashboard capabilities of CloudTrail Lake to boost visibility and evaluation throughout your AWS environments.

The primary one is a Highlights dashboard that gives you with an easy-to-view abstract of the info captured in your CloudTrail Lake administration and knowledge occasions saved in occasion knowledge shops. This dashboard makes it simpler to rapidly establish and perceive necessary insights, akin to the highest failed API calls, traits in failed login makes an attempt, and spikes in useful resource creation. It surfaces any anomalies or uncommon traits within the knowledge.

I’m going to the AWS CloudTrail console and select Dashboard below Lake within the navigation pane to take a look at the Highlights dashboard. First, I allow Highlights dashboard by selecting Agree and allow Highlights.

I try the Highlights dashboard as soon as it populates with knowledge.

The second addition to the brand new dashboard capabilities is a collection of 14 pre-built dashboards. These dashboards are designed for various personas and use instances. For instance, the security-focused dashboards aid you to trace and analyze key safety indicators, akin to high entry denied occasions, failed console login makes an attempt, and customers who’ve disabled multi-factor authentication (MFA). There are additionally pre-built dashboards for operational monitoring, highlighting traits in errors and availability points, akin to high APIs with throttling errors and high customers with errors. You may also use the dashboards centered on particular AWS providers akin to Amazon EC2 and Amazon DynamoDB, which aid you establish safety dangers or operational issues inside these specific service environments.

You may create your personal dashboards and optionally set schedules for refreshing them. This degree of customization helps you tailor the CloudTrail Lake evaluation capabilities to your exact monitoring and investigative wants throughout your AWS environments.

I change to the Managed and customized dashboards to watch the customized and pre-built dashboards.

I select IAM exercise dashboard pre-built dashboard to watch total IAM exercise. You may select Save as new dashboard to customise this dashboard.

To create a customized dashboard from scratch, I’m going to Dashboard below Lake within the navigation pane and select Construct my very own dashboard. I enter a reputation within the Enter a reputation for the dashboard area and select occasion knowledge shops below Permissions, to visualise the occasions. Subsequent, I select Create dashboard.

Now, I can add widgets to my dashboard. You’ve got the flexibleness to customise your dashboards in a number of methods. You may choose from a listing of pre-built pattern widgets utilizing Add pattern widget, or you may create your personal customized widgets utilizing Create new widget. For every widget, you may select the kind of visualization you like, akin to a line graph, bar graph, or different choices to greatest symbolize your knowledge.

Now accessible
The brand new options in AWS CloudTrail Lake symbolize a significant development in offering a complete audit logging and evaluation answer. These enhancements present the flexibility to realize extra profound understanding and conduct investigations extra quickly, aiding with extra preventative monitoring and quicker incident dealing with throughout your complete AWS environments.

Now you can begin utilizing generative AI–powered pure language question era in CloudTrail Lake in US East (N. Virginia), US West (Oregon), Asia Pacific (Mumbai), Asia Pacific (Sydney), Asia Pacific (Tokyo), Canada (Central), and Europe (London) AWS Areas.

CloudTrail Lake generative AI–powered question outcomes summarization functionality is out there in preview in US East (N. Virginia), US West (Oregon), and Asia Pacific (Tokyo) Areas.

Enhanced filtering choices, cross-account sharing of occasion knowledge shops and dashboards can be found in all of the Areas the place CloudTrail Lake is out there, except generative AI–powered summarization characteristic on the Highlights dashboard being accessible solely in US East (N. Virginia), US West (Oregon), and Asia Pacific (Tokyo) Areas.

Working queries will incur CloudTrail Lake question expenses. For extra particulars on pricing, go to AWS CloudTrail pricing.

— Esra