<?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 mahammad thahif on Medium]]></title>
        <description><![CDATA[Stories by mahammad thahif on Medium]]></description>
        <link>https://medium.com/@thahif.86?source=rss-2b785e2c9ff6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*6_ZiMp3zHhL_FC1Z.jpg</url>
            <title>Stories by mahammad thahif on Medium</title>
            <link>https://medium.com/@thahif.86?source=rss-2b785e2c9ff6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 07:23:28 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@thahif.86/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[AWS-Azure:B2B App Migration]]></title>
            <link>https://medium.com/@thahif.86/aws-azure-app-migrations-lessons-learnt-802aef1460ce?source=rss-2b785e2c9ff6------2</link>
            <guid isPermaLink="false">https://medium.com/p/802aef1460ce</guid>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[app-migration]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[refactoring]]></category>
            <category><![CDATA[app-modernization]]></category>
            <dc:creator><![CDATA[mahammad thahif]]></dc:creator>
            <pubDate>Tue, 01 Aug 2023 03:55:05 GMT</pubDate>
            <atom:updated>2023-08-14T06:53:32.271Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/342/1*-8v-tRthUMHbZV5VC7VBaQ.png" /></figure><p>Companies have start adopting cloud technologies more than ever, with flexibility in choosing cloud providers and various services available, more and more organizations are migrating their workloads to cloud.</p><p>As an organizational strategy , companies have moved one step further and envisioned a multi cloud destination for their workloads. Often they align different departments to use a particular cloud provider, sometime an application can span 2 cloud providers! It all depends on an org decision, which application should be moved to which cloud provider and why.</p><p>Let’s look at a scenario, where we migrated a B2B application from AWS to Azure.</p><p>Customer’s B2B portal, hosted completely on AWS , need to be migrated to Azure as is. Though this looks like lift and shift migration, under the hood there is more to it and involves lot of code refactoring as the cloud provider is changing and the way services are consumed through the code are changing.</p><p>Let’s look at the some of the key rationale behind this application migration</p><ul><li>Standardize operational procedures around security, DevOps</li><li>Consolidate to a single cloud provider</li><li>Eliminate unused components in the existing architecture</li></ul><p><strong>Existing AWS Architecture</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/771/1*G4dpHmP5EPzRyBrfhSi0qw.jpeg" /></figure><ul><li>The app is hosted in single region with multi AZ.</li><li>The web and application layer was hosted on IIS web service in an EC2 instance behind an application load balancer (ALB). This is highly available setup with 2 EC2 sitting in 2 different availability zones.</li><li>A separate EC2 server hosting a windows service , which processes notifications from SAP ecosystem.</li><li>A bunch of docker containers are hosted on Elastic Container Service, which are catering to multiple functional aspects of the application.</li><li>The application monitoring is achieved through combination of CloudWatch alarms, CW logs and lambda functions, which in turn calling relevant MS teams webhook endpoints to notify support teams through teams channels.</li><li>There are different SQS queues are setup , which helps in decoupled architecture. These queues are populated with events from SAP ecosystem from on-premise and these are processed by fleet containers hosted on ECS.</li><li>The AuthN/AuthZ is forms based authentication, handled within the application.</li><li>SES being used for email communication and SMS capability.</li><li>A mirrored SQL RDS is used as a backend database.</li><li>The application has multiple integrations, both internal and external. All the external integrations are securely accessed over https.</li><li>Multiple S3 buckets are used for various object storage requirements.</li><li>Redshift is being used as a data warehousing solution, coupled with power BI for visualization.</li><li>Git actions &amp; Octopus were used for CICD requirement.</li><li>Octopus library feature was heavily used for variable mapping for different environment during the deployment phase.</li><li>The on-premise resources were consumed over a AWS direct connect service.</li><li>Terraform was used as cloud agnostic IaC tool.</li><li>Cypress automation test suite was used to perform system ,SIT, regression ,performance testing and UAT</li></ul><p><strong>Azure Target Architecture</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/542/1*LOw74OlrooQKBr7MBticIw.jpeg" /></figure><p><strong>Design Principles</strong></p><p>The target architecture is built based on below listed design principles, which aligns with overall business objectives of the customer.</p><ul><li>Migrate every component from AWS to Azure</li><li>Use PaaS services over IaaS wherever possible.</li><li>Build secure, scalable and cost effective solution.</li><li>Follow zero trust model.</li></ul><p><strong>Tech Stack</strong></p><p>The application technology stack looks like as below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/476/1*Et-bVr8-B6dk_KhelNXVNg.png" /></figure><p><strong>Technology Mapping</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/630/1*9udX2Dyn1e2Qj2o41Pdbkg.png" /></figure><p>One of the most imperative part of this exercise was mapping of AWS services to Azure equivalent services. Though for most of the services we do find equivalent services in different cloud providers, some of them differ the way they’re consumed and minute details surrounding them.</p><p><strong>Key Challenges</strong></p><p>Like any other cloud migrations, we had some key challenges. Let’s look at what are those and how did we overcome it!</p><ul><li>24TBs of S3 objects to be migrated to Azure blob storage.</li><li>Migrating parameter store key-value from AWS SSM to Azure equivalent service.</li><li>Migrating custom CloudWatch math based metric alarms to Azure monitor alerts.</li><li>Application variable replacement through Octopus CD tool</li><li>Code merges from multiple inflight projects.</li><li>Eliminate AWS SNS like dependency and integrate Azure monitor with MS Teams directly.</li><li>Change application logging from AWS CloudWatch logs to Azure App insights.</li><li>Refactor lambda functions to Azure functions. Consolidate multiple runtime to single runtime or achieve the same feat using logic apps.</li></ul><p><strong>How did we SOLVE??</strong></p><ul><li>For S3 data transfer, we used AzCopy tool- A Microsoft freeware which supports S3 as a data source. This tool is powerful, easy to use and is free.</li><li>Since there is no like to like match for AWS SSM parameter store on Azure, these key value pairs were exported and migrated to Azure SQL DB.</li><li>Application variables getting loaded from Octopus were replaced by combination of GitHub secrets and webapp’s app setting features.</li><li>Log analytics KQL based custom queries were developed to build CloudWatch equivalent alerts on Azure.</li><li>Logic App’s MS Teams connector was used to send teams notifications to various teams channels per environment.</li><li>Regular cadence was setup with various teams to collaborate and manage code conflicts/merges.</li><li>Entire code base was re-written to replace reference to S3 with Azure blob storage</li><li>Similarly, code was refactored to replace SQS with service bus, application logging CW logs with app insights.</li><li>Most of the lambda function requirements were achieved through logic apps connector , some were hosted as is on Azure functions. AWS EventBridge scheduling was replaced by logic app scheduling capability.</li></ul><p>Whilst, these are major challenges ,we had overcome other minor issues too during the course of the migration.</p><p><strong>Implementation Approach</strong></p><p>In subsequent sections we shall look at how did we approach implementation of each layer of the application.</p><p><strong>Web Layer</strong></p><p>The Angular frontend was hosted on Azure WebApp behind an application gateway coupled with Azure Web Application Firewall(WAF). All the PaaS services are deployed with private IPs and are not directly accessible over the public network thus making it secure.</p><p><strong>Application Layer</strong></p><p>Application layer too was hosted on Azure WebApp in private mode. The webapp was vnet ingested to control the outbound IP communication. Access restrictions were set ON to allow only few selected services to communicate. The traffic was routed to a firewall solution placed in HUB subscription using a route table.</p><p>The windows service which were hosted on separate EC2 instance was converted to recurring WebJobs.</p><p>A single app service plan(Windows) was used to host all the lower environment’s webapps and function apps.</p><p><strong>Containers</strong></p><p>Azure Kubernetes cluster was used to host docker containers which were previously deployed on AWS ECS cluster.</p><p>Azure Container Registry (ACR) was used to store the docker images.</p><p>An NGINX ingress controller was deployed to expose the underlying services and redirect external traffic to various pods.</p><p><strong>Database layer</strong></p><p>A SQL elastic pool tier was chosen to host all lower environment’s databases. Azure SQL was selected for production use.</p><p>A standard BACPAC import/export method was adopted to move SQL data from AWS to Azure.</p><p><strong>DevOps</strong></p><p>GitHub Actions is used as CICD platform to automate build and deploy pipelines. Deployments yaml are logically separated , one for each type as below.</p><ul><li>WebApp deployment</li><li>WebJob deployment</li><li>Container Deployment</li><li>Function App Deployment</li></ul><p>Each pipeline follows build, resource creation and app deploy model. Pipelines are auto triggered based on changes to TF variable file.</p><p>Container pipelines are triggered based on changes to variable or manifest file.</p><p><strong>Branching strategy</strong></p><p>All the code changes for migration activities were performed under a separate branch namely “Azuremigration”. Individual service level developments were maintained separately.</p><p>Ex: One branch for each service such as service bus, azure blob storage development were created and owned by different developers. After the each feature development, this branch was merged with Azuremigration branch.</p><p>A well-structured Pull Request(PR) process with multiple reviewers and approvers were followed during branch merges and to handle code conflicts.</p><p><strong>Test Strategy</strong></p><p>A comprehensive testing strategy was developed to perform a thorough testing of the migrated application. Below listed testing types were performed post migration.</p><ul><li>ST</li><li>SIT</li><li>Regression</li><li>Performance</li><li>UAT</li></ul><p>Each test phase was properly scoped with entry ,exit criteria and test coverage. Used Jira for effective defect management.</p><p><strong>Pro Tips!</strong></p><p>Use cloud agnostic naming conventions for queues, storage containers etc. As we change the cloud providers it becomes extremely tedious to change the reference to new service(s) if the code is tightly coupled to it.</p><ul><li>Use cloud agnostic IaC tools for resource deployment, this helps in reducing the learning curve incase if you were ever to change the cloud provider.</li><li>Use standard naming and tagging conventions for cloud workloads for easy identifying the service, mapping it to right cost center etc.</li><li>This also helps in quickly building IaC codes just by replacing the environment key words.</li><li>Use WebApp’s deployment slot options to achieve zero downtime based deployments.</li><li>Use the migration activity as an opportunity to revisit some of the unused services in your architecture.</li></ul><ol><li>Ex: we were able to successfully eliminate AWS CloudFront CDN as app was mainly used in ANZ region alone and very minimal traffic from rest of the world.</li><li>Unused S3 buckets were kept out of migration.</li></ol><ul><li>System Managed Identity was extensively used wherever possible instead of traditional connection string based integration.</li><li>Avoid complex S3 bucket policies and instead use Azure RBAC assignment at storage account or container level.</li><li>There is no direct migration path for AWS CloudSearch Index to Azure cognitive search index. Index need to be re-built from scratch. Take this into account while considering timeline for implementing search capability.</li><li>Prepare a detailed test strategy document which includes coverage, scope, tools used and timelines for the testing activities.</li><li>Use defect and risk tracker for effective management and tracking purpose.</li></ul><p><strong>WARNING!</strong></p><p>If you have made this far, you might have noticed we haven’t considered replacement for SES and RedShift, that’s because these 2 services are heavily customized ,due to schedule constraints they were retained AS-IS on AWS world.</p><p><strong>Conclusion</strong></p><p>Cross platform migrations are always complex, time consuming and error prone. Aligning resources who understands both the cloud providers is another challenge one need to keep in mind. Nevertheless this was a challenging exercise involving multiple teams and stakeholders.</p><p>I am sure this was an interesting read and worth the time.Hope it was helpful too .Do drop your comments and thoughts.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=802aef1460ce" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Azure APIM and Application Gateway Integration]]></title>
            <link>https://medium.com/@thahif.86/azure-apim-and-application-gateway-integration-9039378ba197?source=rss-2b785e2c9ff6------2</link>
            <guid isPermaLink="false">https://medium.com/p/9039378ba197</guid>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[api-gateway]]></category>
            <category><![CDATA[apim]]></category>
            <category><![CDATA[azure-application-gateway]]></category>
            <category><![CDATA[api-management]]></category>
            <dc:creator><![CDATA[mahammad thahif]]></dc:creator>
            <pubDate>Mon, 12 Jun 2023 05:34:56 GMT</pubDate>
            <atom:updated>2023-06-12T05:34:56.339Z</atom:updated>
            <content:encoded><![CDATA[<p>Of late, I happened to work on a customer project for setting up a cloud based integration platform. The customer’s existing gateway solutions TIBCO, CA were reaching end of life and license had to be renewed. As part of their long term cloud strategy, they wanted to adopt a modern cloud based integration solution which is secure, scalable and allows cost segregation amongst the different departments of the organization</p><p>The obvious choice was to choose Azure APIM — API management service which provides multiple benefits such as</p><ul><li>Abstract backend architecture diversity and complexity from API consumers</li><li>Securely expose services hosted on and outside of Azure as APIs</li><li>Protect, accelerate, and observe APIs</li><li>Enable API discovery and consumption by internal and external users</li></ul><p>Before we dive in, let’s look at some of the key requirements of the project.</p><ol><li>Design a secure, scalable integration platform that should cater to both internal &amp; external consumers</li><li>The APIM service should not be publicly exposed</li><li>The APIM should allow traffic originating only from a particular region/country</li><li>The communication should be end to end encrypted</li><li>Expose developer portal to only internal consumers</li><li>Enable a chargeback mechanism when the APIs are being used by multiple departments/agencies</li><li>NO DR requirement</li></ol><p><strong>Architecture</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/658/1*_oJ3KyxDcbRjT9U4mt_2sA.png" /></figure><p>To accommodate all of these requirements, we choose Azure APIM combination with application gateway as a front end. The app GW would act as a reverse proxy , protecting the APIM which will be deployed in an internal mode. The App GW is also attached with WAF (v2) instance to accommodate other aspects of the design requirements ,such as Geo filtering and to protect against common OWASP vulnerabilities.</p><p><strong>Implementation</strong></p><p>Create an application gateway in a separate subnet, ensure enough CIDR range to accommodate scaling requirements.</p><p>Ensure below key points while deploying the App GW.</p><ul><li>Since a single app gateway is used to terminate both internal and external traffic, it is essential to create 2 listeners.</li></ul><p>Public-listener — Public IP as the frontend IP</p><p>Pvt-listener — Private IP as the frontend IP</p><p>One important factor to keep in mind, the SSL certificate that needs to be uploaded to listener configuration should be properly created in pfx format. This can be placed in keyvault or manually uploaded from the local machine.</p><p>Whether the certificate was procured from an external vendor or self-signed certificates, make sure it contains data from actual domain pfx certificate as well as from intermediatory certificate.</p><p>In short, we have to export both domain certificate and intermediatory certificates in pfx format, combine them and then upload to listener configuration.</p><ul><li>Create 2 backend pools, one with custom domain name of API GW and another one with custom domain of the developer portal</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/613/1*of-_o2j7cff4wkumzaWZ_g.png" /></figure><ul><li>Create 2 backend settings , one for the gateway and another for the developer portal .Ensure to make it as HTTPS.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/789/1*kACKKFbBo4by57hPNRAY3Q.png" /></figure><ul><li>Create 2 custom probes. Note both Path and response code match parameters.</li></ul><p>Test the probes before proceeding.</p><p><strong>Gateway custom probe properties</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/617/1*9l-2A0I3V8XZOL48ksrkNA.png" /></figure><p><strong>Portal custom probe properties</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/637/1*G67YC6cLJs3IvAB5ZtpYhw.png" /></figure><ul><li>Create <strong>3 Rules</strong> combining all the above settings.</li></ul><p>1 rule for combining public listener, GW backend pool, GW backend setting and custom GW probe.</p><p>1 rule for combining pvt listener, GW backend pool, GW backend setting and custom GW probe.</p><p>1 rule for combining pvt listener, dev portal backend pool, dev portal backend setting and custom dev portal probe.</p><ul><li>When same port is used (443) for both pvt and public listeners, some <a href="https://learn.microsoft.com/en-us/azure/application-gateway/configuration-listeners#frontend-port">NSG rules</a> need to be created.</li><li>Configure app GW specific <a href="https://learn.microsoft.com/en-us/azure/application-gateway/configuration-infrastructure#network-security-groups">NSG rules</a> for the proper functioning of the application gateway</li></ul><p>As a next step, let’s look at the <strong>APIM deployment</strong> key points.</p><p>Deploy APIM in an <strong>internal </strong>mode with appropriate SKU as decided.</p><p>Create custom domain for the gateway and portal URLs.</p><p>Create APIM specific <a href="https://learn.microsoft.com/en-us/azure/api-management/api-management-using-with-internal-vnet?tabs=stv2#configure-nsg-rules">NSG rules</a> for <strong>APIM</strong> to work properly.</p><p>Procure an SSL certificate for your API endpoint. Make sure to procure either a wildcard or SAN certificate to accommodate all the URLs.</p><p>Ex: <a href="https://yourpublicapi.yourdomain.com/">https://yourpublicapi.yourdomain.com</a> — publicly exposed API endpoint pointing to app GW public IP</p><p><a href="https://prodapigw.yourdomain.com/">https://prodapigw.yourdomain.com</a> — API GW custom domain URL mapped to GW URL</p><p><a href="https://proddevportal.yourdomain.com/">https://proddevportal.yourdomain.com</a> — APIM custom domain mapped to developer portal</p><p>This completes the integration of APIM with app gateway from networking perspective. As a next step, start creating APIs with appropriate policies and authorizations.</p><p><strong>DNS Resolution</strong></p><p>Since we have consumers originating from external and internal, it is imperative to get the DNS configured correctly.</p><p>In your external DNS provider (like GoDaaddy) , create an <strong>A record</strong> pointing to <strong>App GW’s public IP.</strong></p><p>In your internal DNS server , create an<strong> A record</strong> pointing to <strong>App GW’s private IP</strong>.</p><p>Create 2 more A records per below to complete the DNS config.</p><p>A record pointing <strong>APIM GW custom UR</strong>L to <strong>pvt IP of the APIM</strong></p><p>A record pointing <strong>APIM developer portal URL</strong> to <strong>pvt IP of the APIM</strong></p><p><strong>WAF Configuration</strong></p><p>Azure Web Application Firewall (WAF) on Azure Application Gateway provides centralized protection of your web applications from common exploits and vulnerabilities.</p><p>Our requirement was to allow traffic emerging from a particular region and enable all the <strong>managed rulesets</strong>.</p><p>Create a custom geo location rule , allowing traffic alone from the required region.</p><p>Initially , enable WAF in <strong>Detection mode</strong> for some days and switch to <strong>prevention mode</strong> to prevent any legitimate request being dropped!</p><p><strong>Note</strong>: If you expose development portal and management endpoint externally through the app GW, make sure to disable <a href="https://learn.microsoft.com/en-us/azure/api-management/api-management-howto-integrate-internal-vnet-appgateway#expose-the-developer-portal-and-management-endpoint-externally-through-application-gateway">mentioned</a> WAF rules for the proper functioning.</p><p><strong>Testing</strong></p><p>Test, Test, Test !</p><p>End to end testing crucial in any implementation, especially in API based setup when requests are initiated from multiple different locations, systems and departments.</p><p>Look at the simple <strong>GET</strong> requests to a backend service served from a 3rd party, notice the API GW URL , which is in turn pointed to app gateway public IP.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/866/1*yJMmgaZqGtZUxrPw7a0hgQ.png" /></figure><p><strong>Additional Tips</strong></p><p>If you use other PaaS service for backend processes such as logic app, service bus or event hubs, enable <strong>private endpoint</strong> as a best practice.</p><p>If your provider works based on IP whitelisting, APIM public IP will be used for making <a href="https://learn.microsoft.com/en-us/azure/api-management/api-management-howto-ip-addresses#ip-addresses-for-outbound-traffic">outbound IP connections</a>.</p><p><strong>Conclusion</strong></p><p>With this long post, we have seen how an APIM can be securely deployed in an internal mode and how App gateway can be front ended to handle the API requests. The app gateway public and pvt IP acts as the entry point for originating requests and accordingly forwards the requests to backend pools.</p><p>This architecture can be extended to include multiple APIM services as backend with appropriate listener configuration done on the app gateway side.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9039378ba197" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Azure Backup — Ransomware Protected Backups]]></title>
            <link>https://medium.com/@thahif.86/azure-backup-ransomware-protected-backups-75ac82d9492c?source=rss-2b785e2c9ff6------2</link>
            <guid isPermaLink="false">https://medium.com/p/75ac82d9492c</guid>
            <category><![CDATA[cyberattack]]></category>
            <category><![CDATA[ransomware]]></category>
            <category><![CDATA[recovery]]></category>
            <category><![CDATA[azure]]></category>
            <dc:creator><![CDATA[mahammad thahif]]></dc:creator>
            <pubDate>Wed, 07 Jun 2023 07:01:54 GMT</pubDate>
            <atom:updated>2023-06-07T07:01:54.493Z</atom:updated>
            <content:encoded><![CDATA[<h3>Azure Backup — Ransomware Protected Backups</h3><p>Ransomware attacks have become one of the biggest security challenges facing businesses today. A successful ransomware attacks can cripple a business core IT infrastructure, and cause destruction that could have a profound impact on the physical, economic security or safety of a business, in turn impacting the brand value.</p><p>Ransomware attacks target businesses of all types. This requires that all businesses take preventive measures to ensure protection.</p><p><strong>How to Protect</strong></p><p>The best way to prevent falling victim to ransomware is to implement preventive measures and have tools that protect your organization from every step that attackers take to infiltrate your systems</p><p>You should assume that at some point in time you will fall victim to a ransomware attack. One of the most important steps you can take to protect your data and avoid paying a ransom is to have a reliable backup and restore plan for your business-critical information. Since ransomware attackers have invested heavily into neutralizing backup applications and operating system features like volume shadow copy, it is critical to have backups that are inaccessible to a malicious attacker</p><p><strong>Azure backup</strong></p><p>Now , let’s look at how Azure backup helped a customer to eliminate a on-premise based 3rd party cyber recovery software footprint and helped reducing licensing cost.</p><blockquote><em>One of the key aspect Azure backup solution is, the underlying storage used for storing the backup is managed by Microsoft and customer/guest/attacker has no direct access to it even if he/she managed to infiltrate the network.</em></blockquote><p>This is because the storage for backup is managed as a service as part of Recovery Service Vault service.</p><p>Whilst this design solves one part of the problem ,it opens up a new attack surface to protect and a new configuration to safeguard the backup configurations. Now, the question arises who can modify backup objects and where can we do it?? Obviously , the Azure portal ,CLI, powershell etc . . .</p><p><strong>Azure RBAC+ MUA + PIM — deadly combination!</strong></p><p>Azure portal management pane provides interface to manage various backup operations as listed below…</p><ul><li>Disable soft delete</li><li>Disable MUA protection</li><li>Modify backup policy (reduced retention)</li><li>Modify protection (reduced retention)</li><li>Stop protection with delete data</li><li>Change MARS security PIN</li></ul><p>Out of these operations, some are highly critical operations and need to be regulated.</p><p><strong>Multi User Authorization(MUA) to rescue</strong> — Multi User Authorization adds an additional layer of protection to safeguard critical operations on recovery services vaults. MUA settings are managed through a special service called Resource Guard.</p><p>Resource guard with MUA adds additional layer of authorization , hence one must have sufficient permission both at Resource guard and recovery service vault level to perform critical operations. To work as expected, resource guard must be owned by a different user ( IT manager/Security admin) and vault admin should not have contributor access on the same.</p><p>Let’s look at how to enable MUA .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/504/1*XTcaIIRsIvAQUfcLFtstdg.jpeg" /></figure><p><strong>Perform critical operations (without Privileged Identity Management)</strong></p><ol><li>Backup admin requests contributor access on the resource guard service.</li><li>Security admin reviews the requests and grants the access.</li><li>Resource manager checks whether backup admins has sufficient permission to perform critical operations at vault level.</li><li>Backup admin perform the critical operation(s)</li><li>Security admin revokes the permission post the activity.</li></ol><p><strong>Perform critical operations (PIM — Privileged Identity Management)</strong></p><ol><li>Create eligible assignment for the backup admin. This can be either security Admin or AAD global admin.</li><li>Create an approval flow adding correct set of people as approvers of the request.</li><li>Backup admin to activate the assigned role.</li><li>Approvers to either approve/reject the request.</li><li>The Backup admin is informed by email on role activation</li><li>Backup admin performs the critical operation(s)</li><li>Access is revoked based on approval duration setting/manual.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/289/1*gtvj3xXQiFC20HW6huSIKQ.png" /></figure><p>When MUA is combined with PIM, it creates a perfect <strong>Segregation of Duty</strong>(SoD) scenario, thereby ensuring one single person cannot perform any disruptive actions on backed up items.</p><p>A perfect workflow would include 3 hierarchies before someone can perform critical actions on the backup items.</p><ul><li><strong>AAD Global Admin</strong> — Persona which creates “Eligible Assignment” in a given Azure tenant</li><li><strong>Security Admin</strong> — This can be your security manager/admin, business recovery/IT manager who approves the requests</li><li><strong>Backup admin</strong> — Cloud operations team member who supports the environment and performs actual operation(s)</li></ul><p>With this additional level of authorization, even if the domain level credentials are compromised , the threat actor(s) cannot gain access to backup copies which are safely protected by Azure storage and guarded by MUA &amp; resource guard.</p><p>Though this solution doesn’t operate as 3rd party cyber recovery softwares it definitely tick marks every objective of ransomware protection softwares available in the market. So one can really build a ransomware protected backup solution using Azure native services, features which are simple, less complex and easy to use and save 3rd party licensing cost.</p><p><strong>Pro Tips!</strong></p><ul><li>Never ever provide backup admin a contributor access to the resource guard service unless authorized otherwise.</li><li>However, backup admin need to have Reader access on the resource guard service to enable MUA on the vault.</li></ul><p>Please find <a href="https://learn.microsoft.com/en-us/azure/backup/multi-user-authorization-concept?tabs=recovery-services-vault#usage-scenarios">usage scenarios</a> and isolation level provided by each of the scenarios..</p><p><strong>Conclusion</strong></p><p>In this ever changing technology landscape , emergence of generative AI platforms like ChatGPT ,hackers/bad actors are equipped with sophisticated methodologies to disrupt the business day in &amp; day out. Amidst all these chaos, there are some rays of hope to protect the data and business reputation. With the benefits of cloud technology and defined operational processes, best practices one can really safeguard IT infrastructure and data.</p><p><strong>Ref</strong>:</p><p><a href="https://learn.microsoft.com/en-us/azure/security/fundamentals/backup-plan-to-protect-against-ransomware">https://learn.microsoft.com/en-us/azure/security/fundamentals/backup-plan-to-protect-against-ransomware</a> <a href="https://learn.microsoft.com/en-us/azure/backup/multi-user-authorization-concept?tabs=recovery-services-vault">https://learn.microsoft.com/en-us/azure/backup/multi-user-authorization-concept?tabs=recovery-services-vault</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=75ac82d9492c" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>