Processing

Please wait...

Settings

Settings

Goto Application

1. WO2022164419 - DISTRIBUTED SEMANTIC-AWARE CACHE

Publication Number WO/2022/164419
Publication Date 04.08.2022
International Application No. PCT/US2021/015022
International Filing Date 26.01.2021
IPC
G06F 12/084 2016.1
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
FELECTRIC DIGITAL DATA PROCESSING
12Accessing, addressing or allocating within memory systems or architectures
02Addressing or allocation; Relocation
08in hierarchically structured memory systems, e.g. virtual memory systems
0802Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches
0806Multiuser, multiprocessor or multiprocessing cache systems
084with a shared cache
CPC
G06F 3/0613
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
3Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
06Digital input from or digital output to record carriers ; , e.g. RAID, emulated record carriers, networked record carriers
0601Dedicated interfaces to storage systems
0602specifically adapted to achieve a particular effect
061Improving I/O performance
0613in relation to throughput
G06F 3/0659
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
3Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
06Digital input from or digital output to record carriers ; , e.g. RAID, emulated record carriers, networked record carriers
0601Dedicated interfaces to storage systems
0628making use of a particular technique
0655Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
0659Command handling arrangements, e.g. command buffers, queues, command scheduling
G06F 3/0689
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
3Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
06Digital input from or digital output to record carriers ; , e.g. RAID, emulated record carriers, networked record carriers
0601Dedicated interfaces to storage systems
0668adopting a particular infrastructure
0671In-line storage system
0683Plurality of storage devices
0689Disk arrays, e.g. RAID, JBOD
Applicants
  • HUAWEI TECHNOLOGIES CO., LTD. [CN]/[CN]
Inventors
  • GKOUNTOUVAS, Theodoros
  • LEI, Hui
  • TANG, Hongliang
  • TANG, Zhihao
  • WANG, Yong
  • WU, Ning
Agents
  • SCHEER, Bradley W.
  • ARORA, Suneel
  • BEEKMAN, Marvin L.
  • BIANCHI, Timothy E.
  • BLACK, David W.
  • LANG, Allen R.
  • NELSON, Albin J.
  • PERDOK, Monique M.
Priority Data
Publication Language English (en)
Filing Language English (EN)
Designated States
Title
(EN) DISTRIBUTED SEMANTIC-AWARE CACHE
(FR) CACHE DISTRIBUÉ SENSIBLE À LA SÉMANTIQUE
Abstract
(EN) A data analytics system implements a data analytics process on a distributed data analytics platform that shares cached data amongst users. The system receives a semantic-aware store command from a user, translates the DAG plan for executing the semantic-aware store command from the user into a translated DAG plan to be executed by a distributed multi-tiered shared cache, and executes the translated DAG plan including at least a final store command. The results of execution of the translated DAG plan are stored to the distributed multi-tiered shared cache. The execution of the translated DAG plan may be optimized by reviewing cached intermediate data to determine whether a branch of the DAG plan can be pruned by loading data directly, whether a repartition of the data can eliminate partitions or shuffle phases, or whether an ensuing operation can eliminate partitions using data-skipping indices.
(FR) Un système d'analyse de données implémente un processus d'analyse de données sur une plateforme d'analyse de données distribuée qui partage des données mises en cache entre des utilisateurs. Le système reçoit une instruction de stockage sensible à la sémantique d'un utilisateur, traduit le plan DAG pour exécuter l'instruction de stockage sensible à la sémantique de l'utilisateur en un plan DAG à exécuter par un cache partagé multiniveau distribué et exécute le plan DAG traduit comprenant au moins une instruction de stockage final. Les résultats de l'exécution du plan DAG traduit sont stockés dans le cache partagé multiniveau distribué. L'exécution du plan DAG traduit peut être optimisée par l'examen de données intermédiaires mises en cache afin de déterminer si une branche du plan DAG peut être élaguée par chargement direct de données, si une répartition des données peut éliminer des partitions ou des phases de brassage ou si une opération suivante peut éliminer des partitions à l'aide d'indices de saut de données.
Latest bibliographic data on file with the International Bureau