Παρακαλώ χρησιμοποιήστε αυτό το αναγνωριστικό για να παραπέμψετε ή να δημιουργήσετε σύνδεσμο προς αυτό το τεκμήριο:
https://hdl.handle.net/20.500.14279/29967
Τίτλος: | NVOverlay: Enabling efficient and scalable high-frequency snapshotting to NVM | Συγγραφείς: | Wang, Ziqi Choo, Chul Hwan Kozuch, Michael A. Mowry, Todd C. Pekhimenko, Gennady Seshadri, Vivek Skarlatos, Dimitrios |
Major Field of Science: | Engineering and Technology | Field Category: | Civil Engineering | Λέξεις-κλειδιά: | Nonvolatile memory;Sockets;Scalability;Subspace constraints;Memory management;Bandwidth;Debugging | Ημερομηνία Έκδοσης: | 14-Ιου-2021 | Πηγή: | Proceedings of ACM/IEEE 48th Annual International Symposium on Computer Architecture (ISCA), 2021, 14-18 June, Valencia, Spain, pp. 498-511 | Start page: | 498 | End page: | 511 | Conference: | ACM/IEEE 48th Annual International Symposium on Computer Architecture (ISCA) | Περίληψη: | The ability to capture frequent (per millisecond) persistent snapshots to NVM would enable a number of compelling use cases. Unfortunately, existing NVM snapshotting techniques suffer from a combination of persistence barrier stalls, write amplification to NVM, and/or lack of scalability beyond a single socket. In this paper, we present NVOverlay, which is a scalable and efficient technique for capturing frequent persistent snapshots to NVM such that they can be randomly accessed later. NVOverlay uses Coherent Snapshot Tracking to efficiently track changes to memory (since the previous snapshot) across multi-socket parallel systems, and it uses Multi-snapshot NVM Mapping to store these snapshots to NVM while avoiding excessive write amplification. Our experiments demonstrate that NVOverlay successfully hides the overhead of capturing these snapshots while reducing write amplification by 29%-47% compared with state-of-the-art logging-based snapshotting techniques. | URI: | https://hdl.handle.net/20.500.14279/29967 | ISBN: | 9781665433334 | DOI: | 10.1109/ISCA52012.2021.00046 | Rights: | © Copyright IEEE - All rights reserved. Attribution-NonCommercial-ShareAlike 4.0 International |
Type: | Conference Papers | Affiliation: | Carnegie Mellon University Samsung Electronics Intel Labs University of Toronto Microsoft Research India |
Εμφανίζεται στις συλλογές: | Δημοσιεύσεις σε συνέδρια /Conference papers or poster or presentation |
CORE Recommender
SCOPUSTM
Citations
20
4
checked on 14 Μαρ 2024
Page view(s) 20
120
Last Week
0
0
Last month
6
6
checked on 31 Αυγ 2024
Google ScholarTM
Check
Altmetric
Αυτό το τεκμήριο προστατεύεται από άδεια Άδεια Creative Commons