This guide to Object Storage (OBS) is a glossary and compilation of the most important Object Storage terms and acronyms. From Erasure Coding to Identifier Address, we’ve put together helpful definitions for anyone looking to understand Object Storage.
For further learning click the hyperlink on each of the terms.
Active archive storage contains the reference or golden copy of data that may be needed for current and future reference, as well as data that must be retained for regulatory compliance, but does not require the performance envelope associated with enterprise-class primary storage solutions.
Bit rot refers to the degradation of magnetic storage whereby certain bits of data are lost due to demagnetization, physical degradation, or other damage to the media whereby data cannot be reliably retrieved.
Cloud storage is a storage model where logical pools of data are distributed across multiple physical storage systems and possibly locations to deliver users uninterrupted easy access to their data through a standard API set from most any location. Storage may be offered through public cloud (shared hosted) services, private cloud (hosted or onsite) deployments, hybrid cloud, or a combination of the two.
Durability refers to longevity of data or the likelihood that the desired data will be successfully retrieved (see bit rot). Data durability is generally achieved through some sort of disk scrubbings such as BitDynamics™ or similar means of verifying the storage content.
Erasure Coding is a data protection scheme that breaks data into shards (fragments or symbols) that are encoded with parity (redundant data), and then stored across multiple storage media and locations.
Fifteen nines is the 99.9999999999999% likelihood of an outcome. Data durability of fifteen nines implies that only 1 in 10 trillion data retrievals would fail.
Geo-spreading is the distribution of data across multiple physical storage systems located in multiple geographic locations through technology such as BitSpread® in order to provide for uninterrupted system availability in case of loss of an entire data center.
Hierarchy-free refers to the absence of a tree or other parent-child structure, such as the flat relationship between data objects in an object-based storage system, which allows extreme scale.
Identifier address is a globally unique address assigned to an object that enables it to be found within a distributed object store without having to know the physical location of the data.
JPEG-friendly is a storage approach that treats pictures or other media as a single object that can be easily indexed and retrieved typically from an object store.
Key Management is the care and handling of security keys that encrypt or lock data against unauthorized access within a shared data center infrastructure. This is typically achieved through key management software internal to a storage system or through external key management appliances.
Large-scale refers to massive deployment such as those commonly found in large enterprises or commercial services providers. In the storage realm, this typically implies data storage that is measured in tens, hundreds, or more petabytes.
Metadata is data that provides information about other data. In the context of OBS, this is descriptive information associated with an object that significantly enhances the ability index and retrieve information based upon the content’s attributes.
In part II we’ll look at everything from Namespace to Zeta-Architecture: The Object Storage Dictionary from A to Z – Part II