OnCallReadyStart learning

Azure IV: Governance, Storage & Backup: interview questions

The question you are most likely to get for each topic, a model answer, and what else comes up.

How would you set up governance for a new Azure environment with several teams? Mid

A management group tree (platform, prod and non-prod landing zones, a sandbox) with subscriptions per team or workload under it. Roles go to Entra groups at the management group level so they are inherited. Azure Policy at the same level: allowed regions, required owner and cost centre tags (modify to inherit them from the resource group), deny public network access where it matters, backup by deployIfNotExists, rolled out as audit before deny with non-compliance messages and exemptions with expiry dates. CanNotDelete locks on production groups. Budgets with forecast alerts per team and Advisor reviews.

Also asked: What is the difference between Azure RBAC and Azure Policy? · How do you give an external partner temporary access to data in a storage account? · How do you prove that your backups work?

A new engineer joins and a contractor starts next week. What do you set up in Entra? Mid

For the engineer: a member account with a verified-domain UPN and a temporary password they must change, a usage location, then membership of the team's groups. Group-based licensing gives the licences and the groups carry the Azure role assignments, so nothing is granted per person. SSPR is on for the group, so password resets don't come to us. The contractor gets a B2B guest invite (az rest to the invitations endpoint, or the portal) and a group with just the role they need, reviewed regularly.

Also asked: What is the difference between a security group and a Microsoft 365 group? · What licence does self-service password reset need for cloud users? · How do you find all guest users in a tenant?

Learn it: 32.1 Microsoft Entra ID: users, groups, licences, guests and SSPR

How would you organise subscriptions and access for a company with many teams? Mid

A management group tree by environment and purpose (platform, prod and non-prod landing zones, sandbox) with subscriptions per team or workload. Roles are given to Entra groups at the management group so they are inherited, which I can check with az role assignment list --include-inherited. Azure Policy is assigned there too. Resource groups follow lifecycles. Tags for owner, environment and cost centre go on everything, enforced by policy, because tags are not inherited and cost reports depend on them.

Also asked: What does a management group give you that a subscription does not? · What happens to a resource's role assignments when you move it to another resource group? · Are tags inherited from the resource group?

Learn it: 32.4 Management groups, subscriptions, resource groups and tags at scale

A deployment fails with RequestDisallowedByPolicy. How do you handle it? Mid

Read the error: it names the policy assignment (and its scope) and the definition, and evaluationDetails shows which field had which value. Then I fix the request, for example the region or a missing tag, because the guardrail is usually right. If this one resource really needs an exception, I create an exemption scoped to it with a reason and an expiry. I don't delete the assignment or widen notScopes, because that switches the rule off for everyone.

Also asked: What is the difference between Azure Policy and RBAC? · What is a policy initiative? · How do you see which resources are non-compliant with a policy?

Learn it: 32.7 Azure Policy: rules that say no (deny, audit, scopes and exclusions)

You assigned a policy that adds a missing tag. New resources have it, old ones don't. Why, and what do you do? Mid

Policy changes resources only when they are created or updated. Existing ones are just evaluated and marked non-compliant. To fix them I run a remediation task (az policy remediation create) after a compliance scan, and it applies the modify effect as the assignment's managed identity. That identity needs a role on the scope, Contributor or Tag Contributor for tags, or every remediation deployment fails.

Also asked: What does deployIfNotExists do, and what does it need? · How would you make sure every VM in a subscription is backed up? · What is a policy alias?

Learn it: 32.12 Policy that fixes things: modify, deployIfNotExists, remediation, initiatives and your own rules

What is the difference between a CanNotDelete and a ReadOnly lock, and which one do you put on production? Mid

CanNotDelete lets authorised users read and change the resource but not delete it. ReadOnly blocks every control-plane write, including POST actions like listing storage keys or starting a VM. Both are inherited and override RBAC, even Owner, and only Owner or User Access Administrator can manage them. On production groups I use CanNotDelete; ReadOnly breaks normal operations, so only resources that must never change get it.

Also asked: Who can create and delete resource locks? · Do resource locks protect the data in a storage account? · A cleanup job fails with ScopeLocked. How do you find the lock?

Learn it: 32.15 Resource locks: the last guard before delete

How do you stop a team from overspending in Azure? Mid

Tags for owner and cost centre, enforced by policy, so cost analysis shows who spends what. A budget per team or subscription with an 80% actual alert to the leads and a forecasted 100% alert to the team's action group, so they hear about it in week two and not on the invoice. Advisor cost recommendations reviewed weekly: right-size or shut down idle VMs, delete unattached disks, buy reservations or savings plans for the steady baseline.

Also asked: What is the difference between a reservation and a savings plan? · What does Azure Advisor recommend, and how is it calculated? · What is the difference between stopping and deallocating a VM for cost?

Learn it: 32.19 Costs: budgets, cost alerts and Azure Advisor

Which redundancy would you choose for a production storage account, and why? Mid

It depends on what it must survive. ZRS keeps three copies in three availability zones, so a zone outage is invisible with no failover. If it must survive a region loss, GZRS adds an asynchronous copy in the paired region, and RA-GZRS lets me read that copy at any time. The geo copy lags by the last sync time, so a failover can lose recent writes. And redundancy is not backup: deletes replicate too, so soft delete, versioning and Azure Backup still matter.

Also asked: What is the difference between LRS, ZRS and GRS? · What are the access tiers and when would you use Archive? · What do you need to use customer-managed keys with a storage account?

Learn it: 32.21 Storage accounts in depth: performance, redundancy, tiers and encryption

A partner needs to download files from one container for two weeks. How do you give them access? Mid

A service SAS on that container with read and list only, HTTPS only, ideally limited to their IP range, tied to a stored access policy that expires in two weeks. Then I can revoke it early by deleting the policy without rotating the account keys. If they were in our directory I would give them Storage Blob Data Reader on the container instead, or a user delegation SAS if a URL is needed. Never an account SAS or the keys.

Also asked: What is the difference between an account SAS and a user delegation SAS? · How do you restrict a storage account to your virtual network? · How would you rotate storage account keys without downtime?

Learn it: 32.24 Access to storage: keys, SAS, stored access policies, data roles and firewalls

Someone overwrote an important blob. Can you get it back? Mid

Only if versioning was on: the old content is a previous version, and I copy that version back over the blob with az storage blob copy start and the ?versionid= URL. Soft delete alone covers deletes, not overwrites. That is why accounts with important data get versioning plus blob and container soft delete, and a lifecycle rule that deletes old versions after a while so they don't grow forever.

Also asked: What is the difference between blob soft delete and container soft delete? · How would you move logs older than 30 days to a cheaper tier automatically? · What is object replication and what are its prerequisites?

Learn it: 32.29 Protecting blob data: soft delete, versioning, lifecycle and object replication

How do you let users open an Azure file share without handing out the storage key? Mid

Identity-based access: join the account to AD DS or use Microsoft Entra Kerberos, give the team's group a share-level role such as Storage File Data SMB Share Contributor, and keep the NTFS ACLs for folder-level control. Users then mount with their own Kerberos identity. Both layers have to allow the action. With nobody using the key any more I can restrict key access, and share snapshots plus soft delete protect the data.

Also asked: What is the difference between share snapshots and share soft delete? · Which port does SMB use, and why does that matter? · When would you use AzCopy instead of the az storage commands?

Learn it: 32.34 Azure Files, identity-based access, and moving data with AzCopy

How do you know your backups actually work? Mid

I restore them. A scheduled restore drill: restore disks or a new VM into an isolated resource group, boot it, check the app or a file, and record the time it took against our RTO. Plus backup job failure alerts routed to on-call, and a report of items whose last successful backup is older than the backup policy allows. A successful job only proves the backup ran; until a restore has worked, I don't count it as a backup.

Also asked: What is the difference between a Recovery Services vault and a Backup vault? · What restore options does Azure Backup offer for a VM? · What is soft delete in Azure Backup and why does it matter for ransomware?

Learn it: 32.36 Azure Backup: vaults, policies, backup and restore

What is the difference between Azure Backup and Azure Site Recovery? Mid

Backup keeps point-in-time copies for restores: an accidental delete, corruption, ransomware, an audit. It loses hours of data and a restore takes a while. Site Recovery replicates VMs continuously to another region for disaster recovery: RPO of minutes, and you fail over the running service, then re-protect and fail back later. Most production systems need both, and both only count if tested: a restore drill for Backup, a test failover for Site Recovery.

Also asked: What are RPO and RTO, and how do you measure them? · What steps does an Azure-to-Azure failover and failback involve? · What besides the VMs must a disaster recovery plan cover?

Learn it: 32.41 Site Recovery: replicate, test, fail over, re-protect

A backup has been failing for weeks and nobody noticed. What do you change? Mid

Route the built-in Backup Failure and security alerts to on-call with an alert processing rule and an action group, at the subscription so every new vault is covered. Send the vaults' diagnostics to Log Analytics in resource-specific mode and add a log alert on AddonAzureBackupJobs for items with no successful backup in the last day. Then fix the cause (usually the VM agent) and schedule restore tests, because a successful job still doesn't prove the data is usable.

Also asked: How do you protect backups against ransomware? · Which alerts does Azure Backup raise by itself? · How would you report on backup health across many vaults?

Learn it: 32.44 Watching backups: alerts, reports, and a vault attackers can't empty

Practise these answers with flashcards and labs Free, in your browser - a real Ubuntu terminal to try it in, with missions that check your work.