M3 - Lesson 1: Overview of DAS, NAS, and SANs (Direct-Attached Storage…
M3 - Lesson 1: Overview of DAS, NAS, and SANs
Direct-Attached Storage (DAS)
However, because you connect DAS storage to the server physically, the storage is unavailable if the server fails.
Storing data locally on DAS makes data centralization more difficult because the data is on multiple servers. This might make it more complex to back up data and users might find it more difficult to locate the data that they want to find.
DAS is physically connected to the server
Easy to configure
Therefore, DAS is the easiest storage system to deploy and maintain.
Almost all servers provide built-in storage, or direct-attached storage. DAS can include disks that are physically located inside the server or that connect directly with an external array, or disks that connect to the server with a USB cable or an alternative connector.
Network-Attached Storage (NAS)
NAS is storage that connects to a dedicated storage device and is then accessed over the network.
NAS differs from DAS because the storage does not attach directly to each individual server, but rather is accessible across a network to many servers.
NAS storage typically is much larger than DAS.
NAS offers centralized storage at an affordable price.
NAS offers a single location for all critical files, rather than dispersing them on various servers by using DAS.
NAS typically includes Redundant Array of Independent Disks (RAID) for data redundancy.
NAS units are accessible from any operating system. They often have multiple-protocol support and can serve up data via CIFS and NFS simultaneously. For example, Windows and Linux hosts can access a NAS unit simultaneously.
You typically access NAS by using Ethernet protocols, and it relies heavily on the network that is supporting the NAS solution.
NAS commonly is used as a file sharing/storage solution, but you cannot (and should not try to) use it with data-intensive programs such as Microsoft Exchange Server and Microsoft SQL Server.
NAS is affordable for small to mid-size businesses, but provides less performance and might be less reliable than a SAN. For this reason, most large enterprises use SANs rather than NAS.
Storage Area Network (SAN)
Requires specialised skills
Fastest access times
High level redundancy
Storage devices that multiple systems share inherently.
Data transfer directly from device to device without server intervention.
A SAN enables multiple servers to access a storage pool in which any server potentially can access any storage unit. However, because a SAN uses a network, you can use it to connect many different devices and hosts, and to provide access to any connected device from almost anywhere.
Block-level storage vs. file-level storage
It is very flexible. For example, you can use it as an operating system volume, a data volume, or a storage location for shared folders.
It is not tied to a specific operating system or a specific file system. All core operating systems and file systems support it.
Operating systems can start from block-level storage LUNs. This means that your organization can deploy diskless physical servers. In such a scenario, the servers use Fibre Channel or iSCSI HBAs to connect to their boot LUN upon startup.
You can present block-level storage directly to virtual machines to meet high-performance storage needs. In Hyper-V, you can present block-level storage to virtual machines by using a pass-through disk or by using virtual Fibre Channel.
Access to file-level storage occurs over file-sharing protocols only.
File-level storage sits on top of block-level storage and has a file system.
Some applications support file-level storage, but others do not. In Windows Server 2012 R2, Hyper-V began supporting virtual-machine storage in SMB 3.0 shared folders.
File-level storage often is more economical than block-level storage.