Librarian View
LEADER 03918nam a22003493i 4500
001
a13378579
003
SIRSI
005
20220514050002.0
006
m d
007
cr un
008
191014t20192019cau om 000 0 eng d
040
a| CSt
b| eng
e| rda
c| CSt
100
1
a| Park, Seo Jin,
e| author.
?| UNAUTHORIZED
245
1
0
a| Achieving both low latency and strong consistency in large-scale systems /
c| Seo Jin Park.
264
1
a| [Stanford, California] :
b| [Stanford University],
c| 2019.
264
4
c| ©2019
300
a| 1 online resource.
336
a| text
2| rdacontent
337
a| computer
2| rdamedia
338
a| online resource
2| rdacarrier
500
a| Submitted to the Computer Science Department.
502
g| Thesis
b| Ph.D.
c| Stanford University
d| 2019.
520
3
a| Today's datacenter applications demand large-scale and low-latency systems. Unfortunately, consistency mechanisms were not designed in consideration of large-scale and low-latency settings. Most existing consistency mechanisms incur huge penalties on scalability or latency, so many datacenter systems have forgone consistency. As a result, application developers or end-users suffer from unpredictable system behaviors. This dissertation presents two new consistency mechanisms for large-scale and low-latency systems: Reusable Infracture for Linearizability (RIFL) and Consistent Unordered Replication Protocol (CURP). RIFL provides a general-purpose mechanism for converting at-least-once RPC semantics to exactly-once semantics, thereby making it easy to turn non-linearizable operations into linearizable ones. RIFL is designed for large-scale systems and is lightweight enough to be used in low-latency environments. On the RAMCloud storage system, RIFL adds only 0.5 us to the 13.5 us base latency for durable writes, and it can support 1 million clients with minimal performance degradation (~5% latency increase). We also used RIFL to construct a new multi-server transaction mechanism (RIFL-TX) in RAMCloud; RIFL's facilities significantly simplified the transaction implementation. The transaction mechanism can commit simple distributed transactions in about 20 us and it outperforms the H-Store main-memory database system for the TPC-C benchmark. Replication is a must-have for large-scale systems for masking individual server failures and maintaining high availability. However, consistent replication incurs significant overhead, such as doubling the latency of operations. As a result, many large-scale systems have settled for weakly consistent replication. To address this dilemma, I present Consistent Unordered Replication Protocol (CURP), which removes most of the overhead of consistent replication. CURP avoids performance penalties by allowing clients to directly replicate their requests, as long as the requests are commutative. This strategy allows most operations to complete in 1 RTT (the same as an unreplicated system). On RAMCloud, CURP improved write latency by ~2x (14 us -> 7.1 us) and write throughput by 4x. Compared to unreplicated RAMCloud, CURP's latency overhead for 3-way replication is just 1 us (6.1 us vs. 7.1 us).
700
1
a| Ousterhout, John K.
e| degree supervisor.
4| ths
0| http://id.loc.gov/authorities/names/n81041478
=| ^A225075
700
1
a| Mazières, David
q| (David Folkman),
d| 1972-
e| degree committee member.
4| ths
0| http://id.loc.gov/authorities/names/no2010058804
=| ^A2466015
700
1
a| Rosenblum, Mendel
e| degree committee member.
4| ths
0| http://id.loc.gov/authorities/names/n94096153
=| ^A2459819
710
2
a| Stanford University.
b| Computer Science Department.
0| http://id.loc.gov/authorities/names/n84052634
=| ^A419917
596
a| 22
035
a| (OCoLC-M)1125917377
856
4
0
u| http://purl.stanford.edu/xc977hj8024
x| SDR-PURL
x| item
916
a| DATE CATALOGED
b| 20191029
999
a| INTERNET RESOURCE
w| ASIS
c| 1
i| 13378579-1001
l| INTERNET
m| SUL
r| Y
s| Y
t| SUL
u| 10/15/2019
x| E-THESIS