Application-Controlled Paging for a Shared Cache (Extended Abstract)
نویسندگان
چکیده
W e consider a cache shared by several concurrently running application processes and propose a provably efficient application-controlled global strategy for the shared cache. Using future information implicitly in the form of good decisions by application processes, we are able to break through the Hk lower bound on competitive ratio proved for classical paging for a k-sized cache in [FKL+91]. For a size-k cache shared by P application processes that always make good cache replacement decisions, we develop an online application-controlled paging algorithm with a competitive ratio of 2Hp-142. Typically, P is much smaller than k , perhaps by several orders of magnitude. Our competitive ratio improves upon the 2P + 2 competitive ratio achieved by [CFLSda]. W e show for this problem that no online algorithm A can have a competitive ratio better than Hp-1 even i f the application processes aiding A have perfect knowledge of individual request sequences. Our results are with respect to a worst-case interleaving of the individual request sequences of the P applications. W e introduce a notion of fairness in the more realistic situation when application processes do not always make good cache replacement deci"Support was provided in part by Army Research Office grant DAAH04-93-G-0076. +Supported in part by the National Science Foundation under grant CCR-9007851 and by the Air Force Office of Scientific Research grants F49620-92-5-0515 and F49620-94-1-0217. 0272-5428/95 $04.00
منابع مشابه
Application-Controlled Paging for a Shared Cache
We propose a provably efficient application-controlled global strategy for organizing a cache of size k shared among P application processes. Each application has access to information about its own future page requests, and by using that local information along with randomization in the context of a global caching algorithm, we are able to break through the conventional Hk ∼ ln k lower bound o...
متن کاملMinimizing Cache Usage in Paging
Traditional paging models seek algorithms that maximize their performance while using the maximum amount of cache resources available. However, in many applications this resource is shared or its usage involves a cost. In this work we introduce the Minimum Cache Usage problem, which is an extension to the classic paging problem that accounts for the efficient use of cache resources by paging al...
متن کاملParavirtualized Paging
Conceptually, fast server-side page cache storage could dramatically reduce paging I/O. In this workshop extended abstract, we speculate how such a device might be used, then show how it can be implemented virtually in a hypervisor. We then introduce hcache (pronounced “aitch-cash”), our prototype implementation built on the Xen hypervisor and utilized by slightly modified Linux paravirtualized...
متن کاملMemory Management in Symunix II: A Design for Large-Scale Shared Memory Multiprocessors†
While various vendors and independent research groups have adapted UNIX and other operating systems for multiprocessor architectures, relatively little work has been done in anticipation of the software requirements of very large-scale shared memory machines containing thousands of processors. Programming environments for these machines must exploit multi-level memory and cache hierarchies so a...
متن کاملPervasive Theory of Memory
For many aspects of memory theoretical treatment already exists, in particular for: simple cache construction, store buffers and store buffer forwarding, cache coherence protocols, out of order access to memory, segmentation and paging, shared memory data structures (e.g. for locks) as well as for memory models of multi-threaded programming languages. It turns out that we have to unite all of t...
متن کامل