Abstract
Non-volatile main memory (NVRAM) is likely to break the bottleneck caused by data transferring between main memory and extern storage, and fundamentally change the way applications do data persistence. We can build persistent data structures directly on NVRAM. To do this correctly and efficiently, we need the support of persistent allocators guaranteeing failure-consistency. However, existing persistent allocators pay huge persistence overhead for failure-consistency. In this article, we present Zallocator, a write-optimized failure-consistency allocator for NVRAM. Zallocator hardly brings write operations to NVRAM. It keeps all the heap management metadata in DRAM, and rebuild them after a crash. Experimental results show that Zallocator achieves a throughput comparable to state-of-the-art transient allocators, and reduces the average number of NVRAM writes per allocation/deallocation to almost zero.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: ACM Journal on Emerging Technologies in Computing Systems
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.