A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2011; you can also visit the original URL.
The file type is application/pdf.
Filters
The case for RAMClouds
2010
ACM SIGOPS Operating Systems Review
Preserved Fulltext
This paper argues for a new approach to datacenter storage called RAMCloud, where information is kept entirely in DRAM and large-scale systems are created by aggregating the main memories of thousands ...
We believe that RAMClouds can provide durable and available storage with 100-1000x the throughput of disk-based systems and 100-1000x lower access latency. ...
In each case the total requirements are no more than a few terabytes, which would fit in a modestsized RAMCloud. ...
doi:10.1145/1713254.1713276
fatcat:vwlxzupmjncujmruqaes4bulee
The case for RAMCloud
2011
Communications of the ACM
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
A new class of storage called RAMCloud will provide the storage substrate for many future applications. ...
the primary storage location for online information in computer systems. ...
the case for ramcloud memories of commodity servers and uses hundreds or thousands of these servers to create a large-scale storage system. ...
doi:10.1145/1965724.1965751
fatcat:nmp2qlgjfvcivhx3fwegyegvuq
An Empirical Evaluation of How the Network Impacts the Performance and Energy Efficiency in RAMCloud
2017
2017 17th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
Through a study carried on RAMCloud, we focus on two settings: 1) clients are collocated within the same network as the storage servers (with Infiniband interconnects); 2) clients access the servers from ...
We compare and discuss aspects related to scalability and power consumption for these two scenarios which correspond to different deployment models for applications making use of in-memory cloud storage ...
for accessing RAMCloud. ...
doi:10.1109/ccgrid.2017.127
dblp:conf/ccgrid/TalebIAC17
fatcat:sqfmdzvgcncfffylubnc6zgviy
Large-Scale Merging of Histograms using Distributed In-Memory Computing
2015
Journal of Physics, Conference Series
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
The running time for distributed merging depends not only on the overall number of bins but also on the number partial histogram output files. ...
Most high-energy physics analysis jobs are embarrassingly parallel except for the final merging of the output objects, which are typically histograms. ...
Acknowledgments We would like to thank John Ousterhout and the RAMCloud team at Stanford University for hosting one of the authors and for numerous suggestions and stimulating discussions. ...
doi:10.1088/1742-6596/664/9/092003
fatcat:vsghcpr6mja7jivslh5vlxldmq
Fast crash recovery in RAMCloud
2011
Proceedings of the Twenty-Third ACM Symposium on Operating Systems Principles - SOSP '11
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
The system uses a log-structured approach for all its data, in DRAM as well as on disk; this provides high performance both during normal operation and during recovery. ...
RAMCloud uses both data parallelism and pipelining to speed up recovery. • Log-structured storage: RAMCloud uses techniques similar to those from log-structured file systems [21] , not just for information ...
In the configuration used for RAMCloud, the memory overhead for tablet profiles is 0.6% in the worst case (8 levels of bucket array for 8 MB of data). ...
doi:10.1145/2043556.2043560
dblp:conf/sosp/OngaroRSOR11
fatcat:iglpm5pr55eajbwylbjhpebxe4
Characterizing Performance and Energy-Efficiency of the RAMCloud Storage System
2017
2017 IEEE 37th International Conference on Distributed Computing Systems (ICDCS)
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
We reveal that although RAMCloud is scalable for read-only applications, it exhibits non-proportional power consumption. ...
We also find that the current replication scheme implemented in RAMCloud limits the performance and results in high energy consumption. ...
As we already explained in section II, RAMCloud monopolizes one core for its polling mechanism, which results in our case in a 25% CPU usage even when the system is idle. ...
doi:10.1109/icdcs.2017.51
dblp:conf/icdcs/TalebIAC17
fatcat:pa7s2k6v4zgu5oa5ldz6vqqfbi
The RAMCloud Storage System
2015
ACM Transactions on Computer Systems
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2018; you can also visit the original URL.
The file type is application/pdf.
In many cases, DRAM is used as a cache for some other storage system, such as a database; this approach forces developers to manage consistency between the cache and the backing store, and its performance ...
In other cases, DRAM is managed in an application-specific fashion, which provides high performance but at a high complexity cost for developers. ...
This eliminated the need to write code for the "rejoin" case. ...
doi:10.1145/2806887
fatcat:fg3r5yahbjhxhcor6m2w2q6bxy
Implementing linearizability at large scale and low latency
2015
Proceedings of the 25th Symposium on Operating Systems Principles - SOSP '15
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2015; you can also visit the original URL.
The file type is application/pdf.
We have implemented RIFL in the RAMCloud storage system and used it to make basic operations such as writes and atomic increments linearizable; RIFL adds only 530 ns to the 13.5 µs base latency for durable ...
We also used RIFL to construct a new multi-object transaction mechanism in RAMCloud; RIFL's facilities significantly simplified the transaction implementation. ...
The presentation in the paper benefited from comments by the anonymous conference reviewers and our shepherd, Steven Hand. ...
doi:10.1145/2815400.2815416
dblp:conf/sosp/LeePKMO15
fatcat:scbxjbbosrdg7jxcwirz4vrerm
Stateless Network Functions
2015
Proceedings of the 2015 ACM SIGCOMM Workshop on Hot Topics in Middleboxes and Network Function Virtualization - HotMiddlebox '15
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2015; you can also visit the original URL.
The file type is application/pdf.
Our Click-based prototype integrates with RAMCloud; using NAT as an example network function, we demonstrate that we are able to create stateless network functions that maintain the desired performance ...
This state separation is enabling these applications to be more agile and support the so-called continuous deployment model. ...
ACKNOWLEDGEMENTS This work was funded in part by the following grants: NSF NeTS 1320389 and NSF XPS 1337399. ...
doi:10.1145/2785989.2785993
dblp:conf/sigcomm/KablanCHJK15
fatcat:grtdkb4xezemlbjymjjifu6kdm
A SURVEY ON CLOUD STORAGE – IMPACT OF STORAGE ON THE PERFORMANCE
2018
International Journal of Research in Engineering and Technology
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
The cloud computing model allows for pay-by-use and thus avoids usage of personal computers or local servers. ...
It is implemented by a connected network of remote servers, accessed via the internet, for processing, storing and managing data. ...
Guaranteed Write RAMCloud is not the good choice for such requirement as in case of the main memory crash (DRAM crash), data will not be written. ...
doi:10.15623/ijret.2018.0710007
fatcat:q54suv7n6jcanieoqwotohdl7u
Improving Memory Access Performance of In-Memory Key-Value Store Using Data Prefetching Techniques
[chapter]
2015
Lecture Notes in Computer Science
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
In addition, we also demonstrate the overhead of prefetching on power consumption. ...
As memory technologies advance, IMKVs become practical for modern Big Data processing, which include financial services, e-commerce, telecommunication network, etc. ...
For example, the first several bytes of value data in RAMCloud contains the size information. ...
doi:10.1007/978-3-319-23216-4_1
fatcat:5d32fhzperd3xkrmo3n5lpa4oa
Exploiting Commutativity For Practical Fast Replication
[article]
2017
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
This strategy allows most operations to complete in 1 RTT (the same as an unreplicated system). We implemented CURP in the Redis and RAMCloud storage systems. ...
Compared to unreplicated RAMCloud, CURP's latency overhead for 3-way replication is just 0.4 us (6.9 us vs 7.3 us). ...
Read operations in CURP may have to wait for the backup sync of the value being read. In the worst-case, one update can block multiple concurrent reads on the same object. ...
arXiv:1710.09921v1
fatcat:ox5t6b2jmnfi3cy4mvczjwydt4
FluidMem: Memory as a Service for the Datacenter
[article]
2017
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
Recent work has begun to explore memory disaggregation, but suffers limitations including lack of consideration of the complexity of cloud-based deployment, including heterogeneous hardware and APIs for ...
We define the requirements of Memory as a Service and build its implementation in Linux as FluidMem. ...
Additionally, we would like the thank William Mortl, Kannan Subramanian, and Daniel Zurawski for their work on this project. ...
arXiv:1707.07780v1
fatcat:thnnbfklg5bmxgnwg4ngtddoly
OFC
2021
Proceedings of the Sixteenth European Conference on Computer Systems
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Using machine learning models adjusted for typical function input data categories (e.g., multimedia formats), OFC estimates the actual memory resources required by each function invocation and hoards the ...
To mitigate this issue, we introduce OFC, a transparent, vertically and horizontally elastic in-memory caching system for FaaS platforms, distributed over the worker nodes. ...
Acknowledgments We thank our shepherd and the anonymous reviewers for their insightful feedback. Experiments presented in this paper were carried out using the Grid'5000 testbed 5 . ...
doi:10.1145/3447786.3456239
fatcat:jdpn6y6ngbgt5bqmkhtbkojnra
Fiber-based architecture for NFV cloud databases
2017
Proceedings of the VLDB Endowment
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
The scalability of the databases (DBs) underlying the virtual network functions is the cornerstone for reaping the benefits from the NFV transformation. ...
We further show that the fiber-based approach outperforms the approach built using conventional multi-threading and meets the variable deployment needs of the NFV transformation. ...
We observe that the multithreaded model scales up to a few cores: two in the case of inserts and five for updates. ...
doi:10.14778/3137765.3137774
fatcat:tlzqna6ukzdy7f7mfqz3qtsipu
« Previous
Showing results 1 — 15 out of 261 results