User Tools

Site Tools



This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
projects:borg:start [m/d/Y H:i]
projects:borg:start [m/d/Y H:i] (current)
Line 1: Line 1:
 ====== Block reORGanization for Self-optimizing Storage Systems ====== ====== Block reORGanization for Self-optimizing Storage Systems ======
 ===== Authors ===== ===== Authors =====
 +  * [[http://​​~mbhad001/​|Medha Bhadkamkar]]
   * [[http://​​~jguer022/​|Jorge Guerra]]   * [[http://​​~jguer022/​|Jorge Guerra]]
   * [[http://​​~lusec001/​|Luis Useche]]   * [[http://​​~lusec001/​|Luis Useche]]
-  * [[http://​​~mbhad001/​|Medha Bhadkamkar]] +  * Sam Burnett 
-  * [[http://​​~rkoll001/​|Ricardo Koller]]+  * Jason Liptak
   * [[http://​​~raju/​|Raju Rangaswami]]   * [[http://​​~raju/​|Raju Rangaswami]]
 +  * [[http://​​~vagelis/​|Vagelis Hristidis]]
 +===== Abstract =====
 +This paper presents the design, implementation,​ and evaluation of
 +BORG, a self-optimizing storage system that performs ​
 +//automatic block reorganization//​ based on the observed I/O
 +BORG is motivated by three characteristics of I/O workloads:
 +non-uniform access frequency distribution,​ temporal locality, and
 +partial determinism in non-sequential accesses.
 +To achieve its objective, BORG manages a small, dedicated partition
 +on the disk drive, with the goal of servicing a majority of the I/O
 +requests from within this partition with significantly reduced seek
 +and rotational delays.
 +BORG is transparent to the rest of the storage stack, including
 +applications,​ file system(s), and I/O schedulers, thereby requiring no
 +or minimal modification to storage stack implementations.
 +We evaluated a Linux implementation of BORG using several
 +real-world workloads, including individual user desktop environments,​
 +a web-server, a virtual machine monitor, and an SVN server.
 +These experiments comprehensively demonstrate BORG's effectiveness
 +in improving I/O performance and its incurred resource overhead.
-===== Introduction ===== 
 ===== Publications ===== ===== Publications =====
 +  * **BORG: Block-reORGanization for Self-optimizing Storage Systems** {{borg-fast09.pdf|pdf}} {{|ps}}\\ Medha Bhadkamkar, Jorge Guerra, Luis Useche, Sam Burnett†, Jason Liptak, Raju Rangaswami, and Vagelis Hristidis\\ Proceedings of File and Storage Technologies (FAST), February, 2009.
 +===== Code =====
 +The code of BORG is not ready to be released. We hope to have a version ready some time soon. However, BORG has similarities to one of our previous projects -- [[..:​exces:​start|EXCES]]. Several details are the same in both projects and thus the EXCES code can be used as reference.
 +===== Traces =====
 +The traces we use to evaluate BORG are available at: [[resources:​start#​traces]]
projects/borg/start.1212531701.txt.gz · Last modified: m/d/Y H:i by ffarfan