Skip to content
  • Home
  • Videos
  • Clean Restores, Confident Recovery: Recovery Orchestration Live

ON-DEMAND SESSION

Clean Restores, Confident Recovery: Recovery Orchestration Live

Learn how Recovery Studio helps teams plan, validate, and execute clean recovery across operational, disaster, and cyber recovery scenarios.

Video thumbnail

Key Takeaways

Practice Before Incidents 

Rehearse recovery in an isolated clean room without disrupting production. 

Validate Clean Points 

Analyze backups for threats, corruption, encryption, and suspicious changes. 

Orchestrate Every Scenario 

Use one runbook model for operational, disaster, and cyber recovery. 

Recover Complete Applications 

Restore workloads and dependencies together into the right environment. 

Reduce Recovery Drift 

Use practiced runbooks during real events with auditable evidence. 

Restore with Confidence 

Return trusted services without bringing threats back into production. 

About this session

Recovery confidence must be earned before an incident. Recovery plans are only valuable when teams have practiced them end to end and can trust the result. This session shows how runbooks, clean rooms, threat analysis, and orchestration turn recovery from a document into a validated operational capability. 

One Framework, Three Scenarios 

Recovery Studio brings operational recovery, disaster recovery, and cyber recovery together under one runbook model. Teams can recover in place, move workloads out of place, rebuild a site, or use an isolated clean room for testing and forensics. 

Validate clean recovery points before restoring. Threat Scan analyzes protected data for indicators of compromise, corrupted or encrypted data, and malware signals. Layered analysis helps teams identify safer recovery points and avoid reintroducing threats into production. 

Practice Before the Bad Day 

Teams can group applications, define recovery ordering and approvals, rehearse runbooks in an isolated environment, and produce evidence for auditors, regulators, and insurers. 

Recover applications as complete units. Runbooks rebuild infrastructure and recover identity, VMs, files, databases, SaaS applications, and AI agents into the right destination. 

Make Recovery Repeatable 

The same runbook language supports production recovery, migration, disaster recovery, and clean-room investigation, helping reduce drift between testing and real events. 

Related resources

WHITEPAPER

Build Identity Resilience

Learn how to protect, detect, and recover identity infrastructure.

Read whitepaper about Build Identity Resilience
SOLUTION BRIEF

Protect Active Directory

Explore visibility, control, and recovery for identity systems.

Read the brief about Protect Active Directory
DATASHEET

Identity Resilience for Okta

Protect Okta objects, view changes, and restore services.

View datasheet about Identity Resilience for Okta

Frequently Asked Questions

How can teams practice recovery safely?

Teams can group application resources, create runbooks, and recover into an isolated clean room without affecting production. They can validate the result, rehearse the flow on schedule, and produce evidence for auditors, regulators, and insurers.

What makes a recovery point safe to use?

A safe recovery point is supported by analysis of protected backup data. Threat Scan can use anomaly detection, threat-intelligence hashes, signature-based analysis, encryption detection, and partner signals to identify malware, corruption, or other recovery risks.

What does clean recovery help prevent?

Clean recovery helps prevent malware or corrupted data from being reintroduced into production. Teams can select a validated point, recover into an isolated environment for investigation or staging, and promote the application only after the recovery is trusted.

Which workloads can recovery runbooks address?

The session describes recovery for identity, virtual machines, files, databases, SaaS applications, and AI agents. Runbooks also account for infrastructure and application dependencies, so teams can recover a complete application unit into the appropriate environment.

How can teams practice recovery safely?

Teams can group application resources, create runbooks, and recover into an isolated clean room without affecting production. They can validate the result, rehearse the flow on schedule, and produce evidence for auditors, regulators, and insurers.

Transcript

View full transcript

Good morning, everyone, and thanks for joining. I’m Dinesh Reddy, and with me is David Cunningham. We are both part of the product management team at Commvault. I’m sure many of you have a recovery plan documented somewhere. My question to you is: have you actually run that plan end to end in the last quarter, or even in the last six months?

That gap between having a plan and trusting it is what this session is about. Over the next 30 minutes, we’ll show you how Commvault turns recovery from a document on a shelf into something that you plan, validate, and execute with confidence. We’ll show it to you in a demo. Here is how we’ll spend our time together.

We have structured this session as a story in three acts. First, some context.

Resilience spans many themes and tools, and that’s exactly why it’s hard. Act one is a good day. That’s when you plan and rehearse so that the bad day is not a surprise. Act two is every day. You continuously scan your backups so you always know where your clean recovery points are. Act three is a bad day.

That’s when you need to recover, whether it’s from a mistake, a disaster, or a cyberattack. Let’s start with the foundation.

Recovery isn’t one problem. It’s three, and today each one is solved with different tools and sometimes by different teams altogether.

First is operational recovery. This is the day-to-day work: accidental deletion and unexpected corruption.

You need recovery to be fast and complete, with services that scale up when you recover and scale down when you are idle. Second is disaster recovery, for which we have Auto Recovery. This is a business continuity solution. When a site or region goes down, you need lower yours and lower yours. You need any do any portability, and you need one-click failover and failback orchestration.

Third is cyber recovery, for which we have Cleanroom Recovery and Threat Scan. This is recovery after a malware attack. That means recovering the control plane itself, scanning for threats, automating the creation of an isolated clean room, and recovering applications as logical groups of workloads. Each of these is strong on its own, but Recovery Studio adds a runbook layer across all of them so you can execute these use cases from one framework.

That brings us to the shift that we are seeing.

Recovery involves more than one scenario, and your platform should handle them all. On a good day, you practice. You run clean-room drills, tests, tabletop exercises, and compliance sign-offs on your own schedule without touching the production environment. On a bad day, someone deletes a folder or a site goes down.

Or perhaps a migration goes sideways. You recover to production, to a new region, or to a different cloud. On a cyber day, the attack threatens the business and destroys trust.

Threat Scan can find a validated clean recovery point, and orchestration using runbooks rebuilds the application in a fresh environment without bringing the threats back.

So Cleanroom Recovery is one use case. Recovery Studio, however, is one platform and one runbook model that covers every recovery scenario. That’s the shift that we are making today. Let’s look at what that actually changes.

Ultimately, cyber resilience is about confidence. When you are attacked, the question is not only how fast you can start recovering. It’s whether you can trust what you recovered. Until now, recovery came in three separate parts.

In-place recovery, Auto Recovery, and Cleanroom Recovery each had their own workflow and interface. None of them had end-to-end orchestration. That meant three tools to learn, three processes to test, and nothing that carried over when the scenario changed.

Now we recover applications as one whole unit. That unit starts with the business applications that you care about.

Order management, payment systems, and customer portals depend on various things, such as identity, DNS, and trust. A single runbook drives all of it in five steps. You group the resources, rebuild the infrastructure, recover the resources in order, validate that the recovery is clean, and then promote the application to production once it is validated.

It’s a zero-touch solution. You can use the same runbook to send the application to different places depending on the scenario. If you want to send the application back to production, you can use an in-place runbook. You can use an out-of-place runbook to send it to a migration site in another cloud or region. Or you can send it to a clean room for isolated testing, forensics, and staging.

This applies to your whole estate, right? Commvault protects identity, VMs, files, databases, SaaS applications, and even AI agents. One engine handles operational recovery, disaster recovery, and cyber recovery. We have packaged all of this as a single solution called Commvault Recovery Studio.

So what is Recovery Studio? Recovery Studio is a unified, runbook-driven recovery platform. It helps you plan, validate, and execute recovery of any protected workload into any environment, with security intelligence built in.

There are four things to remember. First, it’s unified: clean room, body, in-place, and out-of-place recovery all sit in one product and one experience. Second, it’s runbook-driven. The recovery flows live permanently within the platform. Third, it covers the major workloads that Commvault supports, from VMs and databases to cloud-native and SaaS applications. These can be recovered to production, a new region, or an isolated clean-room environment.

Fourth, it’s security-aware. Threat Scan validates recovery points before you commit, so it’s the source of truth for what “clean” means at recovery time. Here is how it fits together.

Start on the left with your production environment. We have VMs, SaaS services, files, databases, and cloud-native workloads, all protected using Commvault Cloud.

Along with the data, we capture the cloud configuration: the network, infrastructure, setup, and everything the applications depend on. The Commvault Cloud control plane drives one runbook framework that recovers into four destinations. The in-place runbook is for operational recovery. It goes back to production from the recovery point that you choose. The out-of-place runbook is for migration testing.

It rebuilds in a new cloud, region, or account. The clean-room runbook is for isolated cyber recovery. It provides an isolated environment for practice and forensic analysis, allowing you to validate the resources before recovering to production. Finally, the disaster recovery runbook provides a full-site rebuild, with cloud configuration and every protected workload orchestrated end to end. These are the same workloads that are in your production environment.

It’s the same runbook language. Only the destination changes. Now let’s walk through the three scenarios, starting with the good day.

Act one is a good day. An outage is the wrong time to find out that your recovery doesn’t work. You need to practice before it happens.

The practice loop has four steps. First, you determine the application group.

That includes everything that makes up your application end to end. Second, you create your runbooks: what to recover, where to recover it, and what the orchestration steps are, such as pauses, recovery ordering, acknowledgments, and approvals. Third, you recover to a clean room. It’s an isolated environment, so there is zero impact on production. This is how you validate your recovery. Fourth, you validate and sign off, giving you auditable proof that you are ready to recover during a bad day.

That gives you three outcomes. You run drills on your schedule with no disruption. You get evidence for auditors, regulators, and insurers. Most importantly, the runbook you practice is the same one that you run during an event, so your test and production recovery never drift apart. Practicing the flow is half the job. The other half is knowing which backup is safe to recover from.

David, over to you to talk about how customers can identify clean points.

Thanks for this. When you take a look at the threat landscape, you understand that cyber threats are continuously evolving.

Threat actors are continuously evolving their tactics, malware payloads, and ransomware payloads. At any given time, as we’re protecting data, those protected snapshots could be impacted by these threats.

There are multiple ways in which that impact could affect your ability to restore.

Number one, you can reinfect your environment if there’s malware hidden within those backups. Number two, you could be recovering data that is corrupted or encrypted by these malware payloads. This is why it’s important to have a solution like Threat Scan that can analyze backup data.

That way, you can analyze your backup data for these threats to drive clean recovery. Threat Scan is a multilayered solution. Multiple signals come together to give customers insights that they can act on for clean recovery. It provides a balance of speed and scale with thoroughness.

It starts with the anomaly detection capabilities in the product. As we protect data, we look at the metadata from those backups and determine whether changes occurred. Those become your early indicators. This is very low-lift: it requires little processing and is very quick, while giving you insight that something changed in your backup.

You may want to inspect it further. The next part of the layered solution is Google Threat Intelligence, and I’ll talk a little about this in the next slide. We’re providing our customers with an industry-best threat intelligence feed to find malware threats within backups rapidly, using hashes provided by Google against that backup data.

When threats are detected, we provide deeper levels of analysis using signature-based engines built into Threat Scan, as well as encryption detection using machine learning. These help customers get a deeper analysis of their data so they know which data is impacted. Lastly, through our partner ecosystem, we use signals from partners such as CrowdStrike and others to provide customers with a 360-degree view of what’s going on in the data they’re protecting.

Think of it as an indicator of attack: suspicious activity on a system that we correlate to the backup. When we combine all of these signals, we provide customers with a comprehensive, layered solution and an understanding of the impact on their backups. Ultimately, the outcome we’re driving is clean recovery.

That clean recovery might be a synthetic recovery, and we’ll talk about that in a little bit. This is clean recovery with minimal rollback. Or it might be a point-in-time recovery, or a recovery for an investigation. All these scenarios are possible because we’re able to analyze the backup data using Threat Scan.

Here’s a high-level overview of how the Threat Scan solution works. It starts with the backup on the top left. As we protect the data, we collect metadata and hashes of the protected data. We use the Google Threat Intelligence feed to search through those hashes. When we detect a match, that means Google has determined that a particular backed-up file is malware or another type of malicious file. That automatically triggers deeper analysis. If no threats are detected, we stop the flow. We did the fast hash lookup using Google Threat Intelligence, and there’s nothing else we need to look at on that system. We also have capabilities that allow customers to schedule periodic deep scans.

As long as the first layer hasn’t detected any threats, no deeper analysis is required. But if threats are detected, we kick off a deep analysis. This examines the data file by file using signature-based engines and our machine learning engines for detecting encryption.

On top of all of this, customers can bring their own IOCs, hashes, and YARA rules to incorporate into the scanning. If SOC teams are looking for specific threats, they can import these into Commvault and use them as part of the scanning for threat hunting or searching for specific threats.

All of this analysis is tracked in our backup index, providing improved visibility in our dashboards. You can look at threats on the dashboard and take action. It allows a fast response, such as clean or safe recovery. By keeping all of this in the backup index, we’re tracking both what’s been protected and the state of that protected data.

For example, we track the threat state: is it a safe file or a suspicious file? This leads us into our synthetic recovery solution.

Synthetic recovery is one of our clean recovery outcomes. One way to perform clean recovery is a point-in-time recovery. You can pick a recovery point from the calendar.

We’ll show you which threats occurred on which recovery points, and you can roll back to a specific point in time.

Our second way of performing clean recovery is synthetic recovery. This is much more granular than a point-in-time recovery. Instead of bringing all your data back to a single point in time, we look at the backup index.

As I mentioned, we look at which files were detected as suspicious or as threats. We go back to the index and try to find the last known good version of those files. All detected malware is automatically quarantined logically. At the index level, we say, “Don’t recover malware,” while anything detected as corrupted or encrypted gets rolled back.

In this illustration, we have several backups that occurred at given points in time. Some of those files were encrypted. When we performed synthetic recovery, we went back to find the last known good version of each file. Why is this important? Our customers told us that clean recovery is one problem they want to solve.

But they also want to limit the amount of data that gets rolled back, and that’s what this does. It limits rollback. What’s powerful here—and I’m going to pass this back to Dinesh—is that all of these recovery options feed into our recovery orchestration. They are part of the recovery orchestration.

Dinesh is going to show us how, once you have this clean recovery data and have completed the analysis, you push that data into your recovery destinations.

Awesome. Thanks, David. We’ve now seen what you do on a good day and every day. Now, act three: a bad day.

Believe me, a bad day comes for all of us. This is where the runbook you practiced and tested on a good day runs against whatever scenario you are facing.

You choose the recovery point. Orchestration handles the rest.

In an operational scenario, someone deletes something by mistake. You recover in place from the latest point. It’s fast and targeted, and it keeps the impact small. In a disaster recovery scenario, a region goes down, or a migration is underway. You rebuild out of place in a new site or another cloud. Then there’s the cyber scenario.

Threat Scan has flagged an anomaly or malware. You validate your application inside the clean room, then recover the clean, validated application to a fresh environment from a clean point. No threat comes along with it. These are three very different scenarios, all handled through one runbook framework. Today, you either have three tools or one tool that handles each scenario.

Recovery Studio handles all three. Okay, enough slides. Let’s see it in action. We’ll show a demo of how this works.

In this demo, we’ll look at how Commvault Recovery Studio helps an e-commerce company called Acme prepare for a cyber event and recover quickly and cleanly from it. First, on a good day, when there is no event, Acme builds and rehearses its recovery. It configures continuous scanning to ensure there is a clean recovery point when needed.

This is Acme Order Hub. It’s healthy. You can see all the records and data. We also have an AI agent helping its users. This is the list of resources that make up Order Hub. As you can see, it’s an extensive set of resources. Acme goes to Commvault Recovery Studio.

It creates a recovery runbook and gives the recovery group a name. Let’s call it Acme Order Group. It creates a runbook, names it Testing Runbook, and submits it. Now the skeleton runbook is created, and it’s time to add resources. First, let’s add a configuration group. The configuration group captures all the Azure cloud configuration information.

That includes the resource groups, subnets, Key Vault, and all of that. Let’s click Restore and specify the restore options we want for this configuration group. If there are any advanced restore options you want to modify, you can do that. Instead of just submitting a restore job for the cloud configuration, let’s add it to the runbook we created.

Now the runbook has cloud configuration as one of the resources that needs to be recovered. Let’s add more resources. Let’s add object storage—Azure File Storage—to it. I’ll pick my instance, click Restore, select the content, and select the restore options. Then, using the same process, let’s add it to the runbook. This way, you can add any workload you want to the runbook.

In my case, I created this runbook with a couple of VMs, a couple of file shares, Active Directory, cloud configuration, and SQL databases. The recovery order matters, right? It’s not just about recovering the resources. It’s about recovering them in the correct priority order. Let’s add a couple of priority steps and rearrange my resources according to priority.

I want cloud configuration to be recovered first, then Active Directory, and then the rest of my resources: VMs, files, and databases. This is the recovery order for my Order Hub application. I’m going to save it, and now my runbook is ready. Whenever I want to test, I can go to this runbook and click Execute Runbook.

Acme wants to rehearse this every month. Instead of a user coming in and executing the runbook each month, Acme goes to Configuration and adds a schedule telling Commvault to execute the runbook every month on XYZ day. This way, even if the user forgets to conduct a recovery test or execute the runbook, Commvault automatically executes it. The user can then download the report.

That was all on a good day, with no event. Acme rehearsed its recovery. Now, a bad day: the admin tries to log in to the Acme Order portal and notices that all the records have been deleted. The database is unreachable. Acme is experiencing a cyber event. Instead of panicking, the admin goes to Threat Scan.

The admin sees critical resources identified with malware and threats. The admin then goes to the recovery runbook that was created and practiced and executes it. During execution, the admin selects a recovery point. There is an option to choose a manual recovery point or synthetic recovery.

By choosing synthetic recovery, the admin can minimize data loss and recover clean copies of the files. Let’s execute synthetic recovery. My recovery has started, following the specified priority order. First, my cloud configuration is recovered, then Active Directory. Once that is complete, the rest of my resources are recovered.

Now I have successfully recovered my Order Hub application using synthetic recovery. When I go back to my Order portal, I can see that all the data is back, including my AI agent. As you can see, the same runbook was rehearsed on a good day and executed on a bad day. Commvault Threat Scan helps identify which recovery point is clean and also helps perform synthetic recovery.

This is how Commvault Recovery Studio, working with Commvault Threat Scan, helps plan, validate, and orchestrate clean recovery during a cyber event.

What you saw was how we plan, validate, and orchestrate recovery end to end.

Instead of separate tools for DR, cyber recovery, and operational recovery, you get one platform for all three.

Instead of choosing recovery points with no security context, you see points validated by Threat Scan when you want to run recovery. Instead of automation that stops a new workload from being recovered, you get a runbook-first approach.

You can recover any workload to any destination. Instead of having runbooks and playbooks outside the environment, those runbooks live within the product. They are practiced, versioned, and auditable. This is the difference that Commvault Recovery Studio provides compared with other solutions.

Thanks, everyone, for joining. If you are interested in the solution, please reach out to your Commvault representatives.

Thanks, everybody.