Free guides, interview Q&As, and job responsibility breakdowns — curated by industry veterans to help you crack MNC interviews
Figure 2: Azure File Sync keeps a local server endpoint in sync with a cloud endpoint through a Sync Group
Azure Files
Definition: A fully managed cloud file share service that can be accessed using the industry-standard SMB and NFS protocols.
Day-to-Day Example: Like a shared network drive at the office, except it is hosted and managed entirely by Azure instead of an on-site server.
File Share
Definition: A single, top-level unit of storage inside a storage account that holds folders and files, similar to a network share.
Day-to-Day Example: Like one shared folder assigned to a specific team, sitting inside a larger cloud storage account.
SMB Protocol
Definition: Server Message Block, a network file-sharing protocol that lets Windows, Linux, and macOS clients mount Azure file shares as a network drive.
Day-to-Day Example: Like the standard front-door key that works for any employee, regardless of which computer they are using.
NFS Protocol
Definition: Network File System, a Linux-native file-sharing protocol supported on Premium Azure file shares over a virtual network.
Day-to-Day Example: Like a private side entrance that only staff already inside the building's network are able to use.
Azure File Sync
Definition: A service that synchronizes on-premises Windows Server file shares with an Azure file share, turning the local server into a fast local cache.
Day-to-Day Example: Like a courier who keeps a branch office's filing cabinet mirrored with the company's central archive.
Cloud Endpoint
Definition: The Azure file share side of a Sync Group, representing where data is stored centrally in the cloud.
Day-to-Day Example: Like the head office's master filing cabinet that every branch ultimately syncs back to.
Server Endpoint
Definition: A specific folder on a registered Windows Server that participates in a Sync Group and syncs with the cloud endpoint.
Day-to-Day Example: Like a branch office's local filing cabinet that mirrors whatever is in the head office's cabinet.
Cloud Tiering
Definition: An Azure File Sync feature that keeps frequently accessed files stored locally in full, while less-used files are represented as pointers to the cloud copy.
Day-to-Day Example: Like keeping this month's paperwork in the office and sending last year's archives to an off-site storage unit, with a note showing where to find them.
Share Snapshot
Definition: A read-only, point-in-time copy of an Azure file share used for backup and recovery of previous file versions.
Day-to-Day Example: Like photographing every page of a filing cabinet on a given day, so you can look back at exactly how it looked then.
Storage Redundancy
Definition: The replication strategy (LRS, ZRS, GRS, GZRS) that determines how many copies of file data Azure keeps, and where.
Day-to-Day Example: Like keeping duplicate copies of important documents in a second building or a second city, in case one location is damaged.
Azure Cosmos DB
Definition: Microsoft's globally distributed, multi-model NoSQL database service built for elastic scale and low-latency access.
Day-to-Day Example: Like a chain of identical warehouses in different countries, each holding the same up-to-date inventory so any nearby customer is served quickly.
Container (Cosmos DB)
Definition: The fundamental unit of scalability in Cosmos DB, holding a set of items that are automatically split across partitions.
Day-to-Day Example: Like a large shipping container that automatically gets divided into smaller crates as more goods arrive.
Item (Cosmos DB)
Definition: A single JSON document stored inside a Cosmos DB container, equivalent to a row in a relational table.
Day-to-Day Example: Like one customer's order form, filed inside the shipping container for that day's orders.
Partition Key
Definition: A property chosen from an item's data that is used to distribute and route items across a container's underlying physical partitions.
Day-to-Day Example: Like sorting mail into bins by postal code so that no single bin ever gets overwhelmed.
Request Unit (RU)
Definition: The currency that measures the compute, memory, and IOPS cost of a single request against Azure Cosmos DB.
Day-to-Day Example: Like a toll booth charging a different number of coins depending on how heavy or complex your vehicle's load is.
Consistency Level
Definition: A setting that defines the trade-off between data accuracy and read/write latency across a Cosmos DB account's distributed replicas.
Day-to-Day Example: Like choosing between reading today's newspaper the moment it's printed versus waiting for the officially confirmed evening edition.
Serverless (Cosmos DB)
Definition: A consumption-based throughput mode in which Cosmos DB charges only for the RUs actually consumed, with nothing provisioned in advance.
Day-to-Day Example: Like paying an electricity bill purely for what you used that month, instead of a fixed monthly subscription.
Autoscale Throughput
Definition: A provisioning mode that automatically scales a container's RU/s up or down within a set range based on real-time usage.
Day-to-Day Example: Like a store automatically opening extra checkout counters during a rush and closing them again once the crowd thins out.
Change Feed
Definition: An ordered, persistent log of every insert and update made to items in a Cosmos DB container, which downstream services can read.
Day-to-Day Example: Like a warehouse's shipping log that a notification system reads to alert the next department the moment new stock arrives.
Multi-Region Writes
Definition: A Cosmos DB capability that lets an application write to the region nearest to it, with changes replicated to every other configured region.
Day-to-Day Example: Like regional branch offices each accepting local customer orders directly, instead of every order having to travel to one head office first.
Figure 3: An Azure Cosmos DB container's items are distributed across partitions using the partition key
Figure 4: The five Azure Cosmos DB consistency levels, from strongest to most eventual
1. SMB vs. NFS
Feature SMB NFS Platform support Windows, Linux, macOS Linux and Linux-based clients only Access Public internet, VPN, or ExpressRoute Virtual network only (no public endpoint) Tier requirement Standard or Premium file shares Premium file shares only Typical use General-purpose file sharing, lift-and-shift apps Linux workloads needing POSIX-style file semantics
2. Standard vs. Premium File Shares
Feature Standard Premium Storage media HDD-based SSD-based Latency Higher, variable Low and consistent Billing model Pay-as-you-go per GB used Provisioned capacity billing Best for General file shares, dev/test High-performance workloads, databases, NFS
3. Azure Files vs. Azure File Sync
Feature Azure Files Azure File Sync What it is The cloud file share service itself A synchronization service on top of Azure Files Where data lives Entirely in the cloud Both on-premises and in the cloud Primary goal Provide a cloud-based SMB/NFS share Keep an on-prem server and the cloud share in sync Local caching Not applicable Yes, via cloud tiering
4. Azure Files vs. Azure Blob Storage
Feature Azure Files Azure Blob Storage Access protocol SMB / NFS (file system semantics) REST/HTTP API (object storage) Typical use Shared drives, lift-and-shift file servers Unstructured data, backups, media, data lakes Mount as a drive Yes No (without a gateway)
5. Manual Throughput vs. Autoscale vs. Serverless (Cosmos DB)
Mode Manual Throughput Autoscale Serverless RU provisioning Fixed RU/s set by admin Scales automatically within a max RU/s None — pay per request Best for Predictable, steady workloads Variable but frequent workloads Intermittent or unpredictable workloads Billing Provisioned RU/s, always billed Billed for highest RU/s used per hour Billed only for RUs consumed
6. Cosmos DB Consistency Levels Compared
Level Data Freshness Typical Latency Impact Strong Always the latest committed write Highest Bounded Staleness Lags by a fixed time/version window High Session Always consistent within the same session Low (default) Consistent Prefix Never sees out-of-order writes Low Eventual May briefly see stale data Lowest
7. Cosmos DB API for NoSQL vs. API for MongoDB
Feature API for NoSQL (Core) API for MongoDB Data model Native JSON documents MongoDB wire-protocol compatible documents Query language Cosmos DB SQL-like syntax MongoDB query language Best for New Cosmos DB-native applications Migrating existing MongoDB applications
8. Account vs. Database vs. Container (Cosmos DB Hierarchy)
Level Represents Example Account The top-level Cosmos DB resource and billing boundary cosmos-retailapp Database A logical grouping of containers SalesDB Container The unit of scalability holding items OrdersContainer
9. LRS vs. ZRS vs. GRS Redundancy
Redundancy Copies Protects Against LRS 3 copies in one datacenter Hardware failure only ZRS 3 copies across availability zones Datacenter failure within a region GRS 6 copies across two regions Regional outage or disaster
10. Cloud Endpoint vs. Server Endpoint
Feature Cloud Endpoint Server Endpoint Location Azure file share Folder on a registered server Role Central, authoritative copy Local, cached copy Count per Sync Group Exactly one One or more
Q1. What is Azure Files, and which protocols does it support?
Answer: Azure Files is a fully managed cloud file share service accessible via the SMB protocol from virtually any OS, and via the NFS protocol from Linux clients on a virtual network.
Q2. What is the difference between the Standard and Premium performance tiers in Azure Files?
Answer: Standard file shares run on HDD-based hardware with pay-as-you-go billing, while Premium file shares run on SSD-based hardware with provisioned capacity billing for lower, more consistent latency.
Q3. What is Azure File Sync, and what problem does it solve?
Answer: Azure File Sync synchronizes an on-premises Windows Server share with an Azure file share, letting an organization keep the performance and compatibility of a local server while centralizing data in the cloud.
Q4. What are the components of an Azure File Sync topology?
Answer: A Storage Sync Service, one or more Sync Groups, exactly one cloud endpoint per Sync Group, and one or more server endpoints on registered servers.
Q5. What is cloud tiering, and why is it useful?
Answer: Cloud tiering keeps only frequently accessed files cached in full on the local server and tiers less-used files to the cloud, which frees local disk space while still presenting a full namespace to users.
Q6. Why does NFS support in Azure Files require a Premium file share and a virtual network?
Answer: NFS 4.1 has no built-in encryption or public endpoint support, so Microsoft only offers it on Premium file shares reachable through a private virtual network for security and performance.
Q7. What is Azure Cosmos DB, and what makes it different from a traditional relational database?
Answer: Azure Cosmos DB is a globally distributed, multi-model NoSQL database service that stores schema-flexible JSON documents and scales horizontally across partitions, unlike a relational database's fixed schema and vertical scaling model.
Q8. What is the resource hierarchy inside an Azure Cosmos DB account?
Answer: An Account contains one or more Databases, each Database contains one or more Containers, and each Container holds a set of Items (JSON documents).
Q9. What role does a partition key play in Azure Cosmos DB?
Answer: The partition key determines how a container's items are distributed across underlying physical partitions, which is central to how Cosmos DB achieves horizontal scale.
Q10. What are Request Units, and why do they matter?
Answer: Request Units (RUs) measure the compute, memory, and IOPS cost of a database operation, and they are the basis for how Cosmos DB throughput is provisioned and billed.
Q11. What are the five consistency levels in Azure Cosmos DB, from strongest to weakest?
Answer: Strong, Bounded Staleness, Session, Consistent Prefix, and Eventual, ordered from the most consistent (and highest latency) to the most available (and lowest latency).
Q12. Which consistency level is the default in Azure Cosmos DB, and why is it popular?
Answer: Session consistency is the default, because it guarantees a user always sees their own writes immediately while still offering low latency for most applications.
Q13. What is the difference between provisioned throughput and serverless mode in Cosmos DB?
Answer: Provisioned throughput (manual or autoscale) reserves a set or scalable amount of RU/s in advance and bills for it regardless of use, while serverless mode has no advance provisioning and bills only for RUs actually consumed.
Q14. What is the Change Feed in Azure Cosmos DB used for?
Answer: The Change Feed exposes an ordered log of inserts and updates to a container's items, which can trigger downstream processing such as an Azure Function for event-driven workflows.
Q15. How does Azure Cosmos DB achieve high availability across regions?
Answer: Azure Cosmos DB can automatically replicate data to any number of Azure regions with a few clicks, and with multi-region writes enabled it offers a 99.999% availability SLA with no application downtime during failover.
Q1. A company wants to retire an aging on-premises file server but still needs staff to map a network drive exactly as before. What Azure service fits best?
Answer: Azure Files fits best, since it can be mounted as an SMB share on Windows machines just like the old file server, with no change to how staff access it.
Q2. A media company has a Linux-based rendering farm that needs a shared file system with POSIX permissions, reachable only from its own virtual network. What should they use?
Answer: They should use an Azure Files Premium share with the NFS protocol, since NFS requires Premium performance and is only reachable over a virtual network, matching both requirements.
Q3. A regional office wants to keep its existing local file server for speed, but the head office wants everything backed up and centrally available in Azure. What should be configured?
Answer: They should deploy Azure File Sync, registering the local server with a Sync Group so its server endpoint stays in sync with a cloud endpoint in an Azure file share.
Q4. A branch server's local disk is nearly full, but most of the files on it are rarely opened. How can the team free up local space without deleting anything?
Answer: They should enable cloud tiering in Azure File Sync, so rarely used files are tiered to the cloud file share while staying visible and accessible from the local server.
Q5. A retail startup needs a database that can handle unpredictable traffic spikes during flash sales without paying for idle capacity the rest of the time. What Cosmos DB option fits?
Answer: Serverless mode fits best, since it charges only for the Request Units actually consumed, with nothing to provision or pay for during quiet periods.
Q6. A global gaming company wants players in Asia, Europe, and North America to all get fast local writes to the same leaderboard data. What Cosmos DB feature should they enable?
Answer: They should enable multi-region writes, so players in each region write to their nearest replica while Cosmos DB replicates the changes to every other configured region.
Q7. A financial application must never show a value that hasn't been fully committed everywhere, even if that means slightly higher latency. Which consistency level should they choose?
Answer: They should choose Strong consistency, since it guarantees every read reflects the most recently committed write across all replicas, at the cost of higher latency.
Q8. An e-commerce team wants new orders to immediately trigger an email notification through an Azure Function as soon as they're written to Cosmos DB. What feature enables this?
Answer: The Change Feed enables this, since it provides an ordered log of new and updated items that can trigger an Azure Function automatically for event-driven processing.
