Explore chapters and articles related to this topic
Big Data
Published in Preston de Guise, Data Protection, 2020
Such options usually require functionality to rebalance in the event of node failure and to generally avoid “hot spots” of data in the distributed filesystem. Distributed filesystems can still benefit from RAID storage however—read performance can be boosted coming from multiple drives, and administrators might prefer RAID rebuild times within individual nodes rather than having entire node rebalancing operations occur in response to a single disk failure.* In fact, some scale-out NAS systems feature compatibility with big data filesystems such as HDFS, allowing HDFS nodes to make use of the scale-out NAS as their actual data storage platform without the need for the multi-copy replication previously mentioned. This can drive greater storage efficiency while also giving access to the rich data services offered by NAS systems. (New storage systems with higher tiers of performance beyond regular SSD—in the tens of millions of input/output operations per second (IOPS)†—have the potential to increase RAID adoption in performance-sensitive big data environments.)
3GPP LTE/LTE-Advanced Radio Access Technologies
Published in Jerry D. Gibson, Mobile Communications Handbook, 2017
In this section, we describe the functions of different protocol layers and their location in the 3GPP LTE/LTE-Advanced protocol structure. Figure 24.2 shows the functional split between the eNB and EPC. In the control-plane, the NAS functional block is used for network attachment, authentication, setting up bearers, and mobility management. All NAS messages are ciphered and integrity protected by the MME and UE. The Radio Resource Control (RRC) sub-layer in the eNB makes handover decisions based on neighbor cell measurements reported by the UE, performs paging of the users over the air-interface, broadcasts system information, controls UE measurement and reporting functions such as the periodicity of Channel Quality Indicator (CQI) reports and further allocates cell-level temporary identifiers to active users. It also executes transfer of UE context from the serving eNB to the target eNB during handover, and performs integrity protection of RRC messages. The RRC sublayer is responsible for setting up and maintenance of radio bearers. Note that RRC sublayer in 3GPP protocol hierarchy is considered as Layer 3 [6,8].
Big Data
Published in Preston de Guise, Data Protection, 2017
Such options usually require functionality to rebalance in the event of node failure and to generally avoid “hot spots” of data in the distributed filesystem. Distributed filesystems can still benefit from RAID storage however—read performance can be boosted coming from multiple drives, and administrators might prefer RAID rebuild times within individual nodes rather than having entire node rebalancing operations occur in response to a single disk failure.* In fact, some scale-out NAS systems feature compatibility with big data filesystems such as HDFS, allowing HDFS nodes to make use of the scale-out NAS as their actual data storage platform without the need for the multi-copy replication previously mentioned. This can drive greater storage efficiency while also giving access to the rich data services offered by NAS systems. (New storage systems with higher tiers of performance beyond regular SSD—in the tens of millions of IOPS† —have the potential to increase RAID adoption in performance-sensitive big data environments.)
iSDS: a self-configurable software-defined storage system for enterprise
Published in Enterprise Information Systems, 2018
Wen-Shyen Eric Chen, Chun-Fang Huang, Ming-Jen Huang
SAN is a block-level data storage solution that uses dedicated networks (Tate et al. 2005). NAS is a file-based network that utilises a general network (Gibson and Van Meter 2000). SAN storage is usually connected by fibre optic cables. On the other hand, a NAS storage is usually connected by network cables. Therefore, data in NAS storage are more constrained than that of SAN due to narrower bandwidths and NAS is commonly used as a file server. Compared to SAN, NAS has the advantage of lower cost. On the other hand, SAN performs better and thus is more suitable for heavy-load workloads. Hyper-converged storage is an emerging storage architecture. It combines storage, computing, networking and virtualisation in one hardware unit (Townsend 2014). It gained popularity because storage can be easily scaled out with more nodes in the environment using virtual machines.