<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Ashutosh Agnihotri on Medium]]></title>
        <description><![CDATA[Stories by Ashutosh Agnihotri on Medium]]></description>
        <link>https://medium.com/@ashutosh.agnihotri?source=rss-15084aac55fc------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*4PxULCCQVyRIL-6MwlU3tQ.jpeg</url>
            <title>Stories by Ashutosh Agnihotri on Medium</title>
            <link>https://medium.com/@ashutosh.agnihotri?source=rss-15084aac55fc------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 12:04:09 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ashutosh.agnihotri/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Enjoy Your Cloud Journey and Keep It Safe]]></title>
            <link>https://medium.com/tech-at-fidelity/enjoy-your-cloud-journey-and-keep-it-safe-8cfceb8c78b2?source=rss-15084aac55fc------2</link>
            <guid isPermaLink="false">https://medium.com/p/8cfceb8c78b2</guid>
            <category><![CDATA[cloud]]></category>
            <category><![CDATA[fidelity-investments]]></category>
            <dc:creator><![CDATA[Ashutosh Agnihotri]]></dc:creator>
            <pubDate>Mon, 05 Oct 2020 14:13:40 GMT</pubDate>
            <atom:updated>2020-10-05T14:41:43.784Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ch6s8KSe6SHZAB7zOeyqnQ.png" /></figure><p>Road trips can be wonderful experiences, and if you are like me, you have probably made some delightful memories on the highway. A road trip consists of many amazing things and each object on the roadside has its own significance. Even the simplest of things like railings! Railings serve to guard you against the dangers that lurk beyond the concrete road. We take them for granted, but they are actually a very important part of the journey; without these important barriers, the voyager might not reach their destination safely.</p><p>At this point, you are probably wondering why I am going on about road trips. Actually, this article is about the cloud journey. I wanted to start the discussion with an analogy that is all about safety. Hopefully, by the end of this article you will be able to draw parallels between the two.</p><p>In this article, I will share ideas that helped us accelerate cloud adoption at Fidelity. We will also look at various service patterns that we enabled for deploying applications in cloud.</p><h4><strong>Service Enablement in Cloud</strong></h4><p>The cloud journey is fascinating, and software developers are eager to take advantage of cloud agility as quickly as possible. The enterprise cloud team’s mission is to provide a frictionless environment to development teams and ensure that applications are always safe and secure. This is facilitated by steering the application deployments in the right direction with the help of appropriate controls and guardrails. The application team will deploy service in cloud using one or a combination of the following patterns:</p><ol><li>Deploy standard patterns</li><li>Deploy non-standard patterns</li><li>Deploy privileged services</li></ol><p>Let us look at the three service enablement patterns for AWS services to understand the challenges with each type and discuss how we enabled them in application pipelines for frictionless deployments.</p><h4><strong>Application Deployments Using Standard Pattern</strong></h4><p>This is the most common and recommended way of deploying applications in cloud. The journey begins by going through a catalog of approved services. A service is considered approved when it has been reviewed by security teams and guidelines for safe usage have been documented. These guidelines consist of the best practices for data protection, access controls, auditing, logging, monitoring and configuration requirements — also known as “Service Guardrails.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*KQS1nEsqHJafm5tqQd7opw.png" /></figure><p>The time and effort spent in defining the guardrails are good investments that every enterprise should consider in their cloud journey. These guardrails prevent applications from drifting away from guidelines set up by corporate security. The process of writing guardrails begins with the review of a new cloud service. Security teams ensure this new service meets the fundamental security requirements, such as:</p><ul><li>Data encryption at rest and in-transit</li><li>Control over the data at all times</li></ul><p>Core functions of guardrails are to provide safety and help accelerate the cloud journey. They are further divided into two controls:</p><p><strong><em>Preventive controls</em></strong> allow application teams to detect violations before deploying resources in cloud. These controls are enforced by governance gates. Going back to our road trip analogy, preventive controls are roadside signboards to warn drivers for hazardous conditions ahead. They may slow down your journey, but they ensure safety.</p><p>The purpose of the<strong> <em>detective controls</em></strong> is to<strong> </strong>ensure services are running safely after they are deployed in cloud. They are analogous to barriers on the road; these guardrails ensure motorists stay on the road while driving.</p><p>These two controls are the cornerstone of application security in cloud. Preventive controls are baked into the application’s deployment pipeline and enforced via governance gate whereas detective controls are seeded at the time of account creation.</p><h4><strong>Non-Standard Pattern Deployments</strong></h4><p>The fact is, not all the applications in your enterprise will be able to finish the cloud journey by staying within the guardrails. For example, an application may require hosting of static content from the S3 bucket. This pattern will require a publicly accessible bucket with a global read permission. Security teams might have set up the guardrails and controls to prevent public access on S3 buckets to protect the sensitive data. This conflict will lead to a friction point and application deployments will be blocked at governance gate due to preventive control violation.</p><p>To solve this problem in Fidelity, we built an exceptions management system. It enables the application deployments to use an unapproved service. To remove the friction completely, the exception management system must provide an easy integration point for application pipelines.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*zzYDDDLb0Z7joMMtGR0VZg.png" /></figure><p>As in your road trip, you take frontage road when driving conditions on highway are hazardous. Those detours will slow you down, but at the same time ensure your safety. Similarly, when a team designs an application that deploys a non-standard (unapproved) service, the deployment will fail the standard checks. These failures will surely introduce friction points in the first implementation of the application. Non-standard patterns require a review from security teams. This manual approval process introduces delays for the first deployment. But once the application teams pass this initial hurdle, deployment is put back on fast track again. Future deployments are frictionless since all controls and gates that the application goes through in its lifetime are aware of the approved exception for the non-standard pattern.</p><h4><strong>Privileged Service Deployment</strong></h4><p>As a part of the best SDLC practices, we expect application developers to have the ability to deploy infrastructure as code, including, but not limited to, IAM and network security. These core services pose too much risk to the firm when not handled correctly. Therefore, core services need custodians to ensure privileged services are always protected in AWS accounts. As in your road trip, you will rent a car by presenting your driver’s license and credit card to car rental company. They will require these documents to ensure you are a safe and trustworthy driver. Likewise, AWS service custodians will let other teams use the protected services after making sure requested change is authorized.</p><p>We quickly realized this requires a fresh look at how we authorize changes in the cloud environment. To solve this problem, we needed a way to allow users from the DevOps team to be authorized to configure protected resources. Therefore, we built a new authorization model specially designed for cloud accounts, which enables:</p><ol><li>Self-Service</li><li>Scalability to all the business partner’s requirements for managing infrastructure</li><li>Easy consumption of APIs that are facilitating privileged service configuration</li></ol><p>As illustrated in the diagram below, a <strong>robust</strong> <strong>authorization model is enabler</strong> for this pattern.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*HAOXYUmqCTK4ZomIh6Fvlw.png" /></figure><p>It took multiple attempts to find the right authorization model that worked at-scale for a large enterprise like Fidelity. Earlier attempted models were based on the Configuration Management DataBase (CMDB). These models enabled smaller business units but quickly became bottleneck for larger business units.</p><p>Some of the challenges we encountered in the early days of our cloud journey were:</p><ul><li>Central infrastructure team responsible for managing changes across multiple products, were blocked due to strict rules around product line boundaries in CMDB.</li><li>Authorization based on change management ticket did not go well with multi-app tenancy in AWS accounts.</li><li>This was a highly opiniated solution for authorization. If you want to takeaway one thing from this article, it would be this one. Never build opiniated solutions!</li></ul><p>These challenges were finally overcome by introducing an abstraction layer for authorization in cloud. This abstraction allowed our business partners to have full control over the process. AWS account owners were able to authorize users by putting them in an enterprise Active Directory (AD) group as per their role, and then associating that AD group to a role for accessing privileged services in AWS account.</p><p>As depicted in the following picture, authorization can be scaled to any number of roles. But it is a good idea to have at least four roles for managing IAM, Network, Perimeter and Access controls:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/515/1*x_usgHKW2g2P6WnzHy6aQg.png" /></figure><p>This authorization model caters to the following two scenarios which cover most of the business units in Fidelity:</p><p><em>Scenario 1</em>: Smaller business units might have one team for managing all functions in AWS accounts for the application development lifecycle (Infrastructure, security and DevOps are functions of same team)</p><p><em>Scenario 2</em>: Larger business units where separation of duties is required between DevOps and Infrastructure teams. For example, the DevOps team may not be allowed to make changes in IAM or Network security, but they should be allowed to access EC2 instances in order to troubleshoot a production issue.</p><p>Role based authorization via AD groups gives account owners full control. In the first scenario, they can put all of the users in one AD group and then assign that AD group to all roles. This will authorize one team and grant access to all of the privileged services in AWS accounts; and in the second scenario they can have four separate AD groups and authorize teams corresponding to their role in the organization.</p><h4><strong>Final Thought</strong></h4><p>Using these deployment patterns for AWS services, Fidelity teams can deploy applications in cloud without any human intervention. Developers deploy with confidence as services are always guarded by controls ensuring the applications are always secured. This ecosystem works well when APIs are considered first class citizens in the enterprise. APIs enable code delivery in production through pipelines regardless of application pattern.</p><p>Disobeying the traffic rules on your road trip could result in life-threatening injuries; similarly, applications deployed in cloud without proper controls and guardrails could result in severe security holes. Stakes are very high in both cases, so never ignore the guardrails and controls put in by security teams. Drive safely and enjoy the rest of your cloud journey!</p><p>#fidelityassociate</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g1x_hJ0lyLW7npMkoUCBFQ.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8cfceb8c78b2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/tech-at-fidelity/enjoy-your-cloud-journey-and-keep-it-safe-8cfceb8c78b2">Enjoy Your Cloud Journey and Keep It Safe</a> was originally published in <a href="https://medium.com/tech-at-fidelity">Tech at Fidelity</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>