Let’s get back to basics and talk about what object storage is, how it compares to traditional file and block storage, and when object storage is the right solution for your data center.
There are 3 main network storage technologies:
- Block: Data is organized through block IDs (e.g., sector number) and can be organized as a structure (called a file system) or an application-specific structure. The client operating system accesses block storage through Fibre Channel or iSCSI or using a direct-attached storage device (DASD). Block storage is ideal for transactional or structured information like file systems, databases, transactional logs, swap space, or for running VMs. It is optimized for block-level performance measured in IOPS (Input/output operations per second). Using traditional file systems on block storage places explicit or practical operational limits on scaling beyond the petabyte range.
- File: Data is accessed as file IDs (server name + directory path + filename) over a shared network and the storage server manages the data on disk. NFS and SMB are the common network protocols used for file access over a network. The storage server or array uses block storage with a local file system to organize these files, and clients only deal with the protocol and the file path. Fixed file attributes like type, size, date created and date modified are stored in the file system. File-based storage is good at shared files and shared directories over a LAN (local area network) or WAN (wide area network). The areas where this kind of network-attached storage (NAS) runs into problems is with the scaling limits of their underlying file system and with their inability to spread workload across multiple file servers.
- Object: Data storage is based on key value addressing (store an object and get a key, just like a car valet giving you a ticket). For object storage, the client or access method is usually an application over HTTP and custom information about the file is stored in its metadata. Object storage is ideal for shared files which can be stored as-is or deleted and for highly scalable, multi-site deployments.
There are a number of challenges that motivate organizations to adopt object storage. Here are a few of the top ones:
- Risk of data loss, data growth and limitations in traditional technologies (which start to stutter at the petabyte capacity range)
- Users expect always on and accessible storage and access from new web-based applications
- Inefficient data silos or locking data into a single location, limiting the ability for reuse and analysis—particularly in big data, life sciences and medical imaging use cases where sharing information can literally be a matter of life or death
Object storage can solve many of today’s storage challenges. Want to learn more? Join us on our BrightTALK channel in 2017, where we will delve into the basics of object storage, talk about object storage for developers, and share customer success stories with organizations that have adopted object storage. Subscribe to Caringo’s BrightTALK channel.
Editor’s note: If you enjoyed this blog, watch the on-demand webinar, Back to Basics: What is Object Storage?