A. Vahdat and T. E. Anderson. Transparent Result Caching. In Proceedings of the USENIX Annual Technical Conference (USENIX '98), New Orleans, Louisiana, June 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....by virtue of manual I O division: all endpoint I O happens via explicit network communication. Yet, we can hardly expect that all valuable applications will be re written for a distributed environment. Ideally, such I O roles would be detected automatically. Such an approach is taken by the TREC [30] system, which deduces program dependencies from I O behavior. We might also reasonably ask the user to provide hints of I O roles to the system without modifying applications directly. A number of file systems take account of the conventional wisdom that quickly deleted data is a significant ....

A. Vahdat and T. Anderson. Transparent result caching. Technical Report CSD-97-974, Computer Science Division, University of California-Berkeley, 1997.

....track the state of the client processes, has also been introduced. The benefits of this approach include improved recovery speed and the ability to hide client failures from servers. It can be combined with other application level proxies such as WAP gateways [2] HTTP proxies for both active [15, 20, 11, 19] and passive [8, 18, 3] content, web clipping proxies [16] Request Request Response Response HTML Script Encoder Decoder GUI Engine Client Server Proxy Request Request Response Response HTML Script Encoder Decoder GUI Engine Client Server Proxy Figure 1. Application level ....

A. Vahdat and T. Anderson. Transparent Result Caching. In Proceedings of the USENIX

....a building block for a scalable consistency system. 4. Uncachable error. Objects are marked uncachable or encounter errors for a number of reasons, some of which might be addressed by more sophisticated cache protocols that support better cache consistency, caching dynamically generated results [31], dynamically replicating servers [33] negative result caching [5] and caching programs along with data [4, 32] We do not address such protocol extensions here. Also, because we are interested in studying the effectiveness of caching strategies, for the remainder of this DEC Berkeley Prodigy ....

A. Vahdat and T. Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, June 1998. To appear.

....[5] 6] and on the Alexandria Digital Library system [7] demonstrated the feasibility and significance of dynamic content caching. A major problem with dynamic content caching is to ensure that strong consistency be maintained between cached pages and underlying data objects. Vahdat et al. [8] and Holmedehl et al. 7] have proposed invalidation techniques using file operation interception or TTL for a class of applications. Iyengar et al. 5] 6] proposes a general approach which lets applications explicitly issue invalidation messages to a cache. While this approach is powerful for ....

....rules [3] 4] which can exploit URL patterns. These studies do not address invalidation of dynamic pages in details. It is more appropriate to cache dynamic pages at sever sites because of their strong consistency requirement and the difficulty to invalidate cached pages on proxies [5] 7] 6] [8]. To invalidate dynamic pages, Holmedehl et al. 7] uses the standard time to live (TTL) technique. Vahdat et al. 8] proposes to intercept file system calls for monitoring source file changes. These techniques are only effective for limited applications. Iyengar et al. 6] 5] propose a ....

[Article contains additional citation context not shown here]

Amin Vahdat and Thomas Anderson, "Transparent Result Caching," in Proc. of 1998 USENIX Technical Conference, 1998.

....[5] 6] and in the Alexandria Digital Library system [7] demonstrated the feasibility and significance of dynamic content caching. A major problem with dynamic content caching is to ensure that strong consistency be maintained between cached pages and underlying data objects. Vahdat et al. [8] and Holmedehl et al. 7] have proposed invalidation techniques using file operation interception or TTL for limited applications. Iyengar et al. 5] 6] proposes a general approach which lets applications explicitly issue invalidation messages to a cache. While this approach is powerful for many ....

....rules [3] 4] which can exploit URL pattens. These studies do not address invalidation of dynamic pages in details. It is more appropriate to cache dynamic pages at sever sites because of their strong consistency requirement and the difficulty to invalidate cached pages on proxies [5] 7] 6] [8]. To invalidate dynamic pages, Holmedehl et al. 7] uses the standard time to live (TTL) technique. Vahdat et al. 8] proposes to intercept file system calls for monitoring source file changes. These techniques are only effective for limited applications. Iyengar et al. 6] 5] propose a ....

[Article contains additional citation context not shown here]

Amin Vahdat and Thomas Anderson, "Transparent Result Caching," in Proc. of 1998 USENIX Technical Conference, 1998.

....and personalized information presentation. There are a number of techniques proposed to speedup dynamic page delivery, for example, efficient OS support [4, 22] server clustering and load balancing [3, 9, 12, 29] content preprocessing [10] and caching dynamic content at server and proxy sites [14, 15, 24, 25]. This paper focuses on caching and invalidating dynamic pages at the server sites. Previous work on Web caching and invalidation such as [7, 13] at proxy servers mainly deals with static pages. Caching dynamic pages at server sites has several advantages: 1) it reduces the server computing ....

....in the IBM Olympics Web server [8, 15] and in the Alexandria Digital Library system [2, 14] demonstrated the feasibility and significance of caching dynamic content. To maintain data consistency for cached dynamic pages, Smith et al. [14] uses the standard time to live (TTL) technique. Vahdat et al. [25] proposes the idea of intercepting file operation related system calls to monitor change of source files and the weakness is that their monitoring is limited to local file systems and it can slow down other operations. For more complicated servers which involve queries on distributed database ....

[Article contains additional citation context not shown here]

A. Vahdat and T. Anderson. Transparent Result Caching. In Proc. of 1998 USENIX Technical Conference, 1998.

....and personalized information presentation. A number of techniques have been proposed to speed up dynamic page delivery, including efficient OS support [5, 24] server clustering and load balancing [2, 11, 15, 31] content preprocessing [12] and dynamic content caching at server and proxy sites [17, 18, 26, 27]. This paper focuses on caching and invalidating dynamic pages at the server sites. Previous work on Web caching and invalidation [8, 16] at proxy servers mainly deals with static pages. Caching dynamic pages at server sites has several advantages: 1) it reduces server computing resource demands; ....

....A major problem for dynamic content caching is to ensure that strong consistency be maintained between cached pages and underlying data objects. Holmedehl et al. 17] uses the standard time to live (TTL) technique which is applicable to limited applications such as online searches. Vahdat et al. [27] proposes 1 the idea of intercepting file operation related system calls to monitor change of source files. The weakness is that system call monitoring is limited to local file systems and can slow down other operations. For more complicated servers which involve queries on distributed database ....

[Article contains additional citation context not shown here]

A. Vahdat and T. Anderson. Transparent Result Caching. In Proc. of 1998 USENIX Technical Conference, 1998.

....not a facility for debugging. In particular, the OS designer should target security and extensible reference monitors as one important application. Other bene ts of generalized interposition mechanisms are documented in papers describing new lesystems [5, 39, 43, 42] transparent result caching [67], emulation of other operating systems [43, 42] transactional software environments [43, 42] and so on. Because these applications can be handled with very similar mechanisms, we argue that one generalized mechanism could handle them all with little additional cost. On the other hand, security ....

Amin Vahdat and Thomas Anderson. Transparent result caching. In Proc. 1998 USENIX Technical Conference, 1998.

....a building block for a scalable consistency system. 4. Uncachable error. Objects are marked uncachable or encounter errors for a number of reasons, some of which might be addressed by more sophisticated cache protocols that support better cache consistency, caching dynamically generated results [31], dynamically replicating servers [33] negative result caching [5] and caching programs along with data [4, 32] We do not address such protocol extensions here. Also, because we are interested in studying the effectiveness of caching strategies, for the remainder of this study, we do not ....

A. Vahdat and T. Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, June 1998. To appear.

.... the number of clients sharing a cache system [33, 54] allowing servers to send new data to cache systems before any nodes in the system request the data, providing better cache consistency [58, 80, 127] to reduce the number of requests marked as uncachable, caching dynamically generated results [113], dynamically replicating servers [115] and negative result caching [83, 19] Since we are interested in studying the effectiveness of caching strategies, for the remainder of this study, we do not include Uncachable or Error requests in our results. 5.2.2.2 Sharing Figure 5.3(a) ....

A. Vahdat and T. Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, June 1998. To appear.

....permit caching of authenticated content and it hinders cache replacement algorithms based on job execution time. In a project currently in progress at UC Berkeley, Vahdat and Anderson propose a framework for monitoring source files, automatically invalidating the result whenever the source changes [16]. An example application that uses the monitoring system is dynamic content caching: they have modified a version of the Apache HTTP server to cache CGI results, and use their monitoring system to invalidate cached entries. Their model does not provide any method for limiting the size of the cache ....

.... need stronger content consistency, we plan to investigate other cache entry invalidation methods in future versions of Swala, for example by receiving invalidation messages from applications [12] or by monitoring the input of the CGI programs whose output is being cached, to detect invalidation [16]. We address cache table consistency with a two level consistency protocol: intra server and inter server consistency. In the following section we describe our cache table consistency protocol. Our intra node consistency protocol protects the internal cache directory against simultaneous updates, ....

A. Vahdat, T. Anderson, "Transparent Result Caching," draft accepted for publication in USENIX '98, http://now.cs.berkeley.edu/WebOS/papers/trec.ps.

.... the number of clients sharing a cache system [10, 18] allowing servers to send new data to cache systems before any nodes in the system request the data, providing better cache consistency [21, 25, 41] to reduce the number of requests marked as uncachable, caching dynamically generated results [36], dynamically replicating servers [37] and negative result caching [27, 5] Since we are interested in studying the effectiveness of caching strategies, for the remainder of this study, we do not include Uncachable or Error requests in our results. 2.2.3 Sharing Figure 3 illustrates the ....

A. Vahdat and T. Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, June 1998. To appear.

....block for a scalable consistency system [10] 4. Uncachable error. Objects are marked uncachable or encounter errors for a number of reasons, some of which might be addressed by more sophisticated cache protocols that support better cache consistency, caching dynamically generated results [41], dynamically replicating servers [43] negative result caching [31, 8] and caching programs along with data [7, 42] We do not address such protocol extensions here. Also, because we are interested in studying the effectiveness of caching strategies, for the remainder of this study, we do not ....

A. Vahdat and T. Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, June 1998. To appear.

....methods in future versions of Swala. This can be implemented by receiving invalidation messages from applications after the model of Iyengar et al. 13] and or by monitoring the input of the CGI programs whose output is being cached, to detect invalidation, as suggested by Vahdat and Anderson [19]. We address cache table Cached yes Return CGI result no Exec CGI and tee results to cache file Fetch from remote cache yes no Create cache space Success Local Create cache entry Return error message Fetch from local cache Success yes yes no no Exec CGI, no caching Success no yes Update ....

A. Vahdat, T. Anderson, "Transparent Result Caching," unpublished.

.... have proposed a bewildering array of new services for mediating between clients and servers, including dynamic redistribution of replicas over the wide area [49] compression and distillation of multimedia content [3, 22] proxy cache extensions such as support for hit counting and dynamic content [11, 48], client customization of web content [54] network address translators (NATs) 20] packet delivery services for mobile hosts [13] and so forth. One approach to deploying such extensions would be to develop a new protocol that closely coordinate browsers, proxies, servers, and the name system to ....

.... [22, 3] compression and delta encoding[39] client or either Consistency verify (unmodified) 10 (ISP) 2 7 [18] 4 13 [5] Server driven consistency [36, 55] server Consistency miss 0 4 [18] delta encoding[39] either Dynamic (cgi or query) 21 (ISP) 0 34 [37] Active cache [11] HPP [17] server TREC [48] server Pragma: no cache 9 (ISP) 5.7 7.2 [26] Hit logging server Active cache [11] function shipping server Server driven consistency [55] server Redirection 3.7 (ISP) Server selection anycast [8, 56] server Table 1: Workload requirements. Numbers are taken from the literature as noted or from ....

Amin Vahdat and Thomas Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, New Orleans, Louisiana, June 1998.

....services. In addition to automatic load balancing and failover among server replicas, examples include dynamic redistribution of replicas over the wide area [66] compression and distillation of multimedia content [28, 3] proxy cache extensions such as support for hit counting and dynamic content [14, 65], client customization of web content [18, 71] network address translators (NATs) 26] packet delivery services for mobile hosts [16] and so forth. To address this need for rapid deployment and extensibility of Internet services, a variety of proposals have been made to support active ....

.... compression and delta encoding[53, 8] client or either Consistency verify (unmodified) 10 (ISP) 2 7 [23] 4 13 [7] Server driven consistency [50, 46, 73] server Consistency miss 0 4 [23] delta encoding[53, 8] either Dynamic (cgi or query) 21 (ISP) 0 34 [51] Active cache [14] HPP [22] server TREC [65] server Pragma: no cache 9 (ISP) 5.7 7.2 [36] Hit logging server Active cache [14] function shipping server Server driven consistency [46, 73] server Redirection 3.7 (ISP) Server selection anycast [74, 11] server Table 1: Workload requirements. Numbers are taken from the literature as noted or ....

Amin Vahdat and Thomas Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, New Orleans, Louisiana, June 1998.

....100 million in less than 5 years [34] To address these problems, we are building WebOS, a framework for supporting geographically distributed, highly available, incrementally scalable, and dynamically reconfiguring applications. WebOS includes mechanisms for global naming [48] persistent storage [43, 42], remote process execution, resource management, authentication and security [4] We use WebOS to demonstrate the synergy of these services in simplifying the development of wide area distributed applications and in providing more efficient global resource utilization. The WebOS framework enables ....

....present an overview of the WebOS effort and a snapshot of the current status of our implementation. WebOS is composed of a number of different components. In the space allowed, we can only briefly describe the main ideas behind individual components, each of which is described in detail elsewhere [48, 43, 4, 42]. We have completed prototypes of each component and are now working to integrate them. The end goal is to provide a platform for the development and deployment for wide area applications. Toward this end, we demonstrate an extensible mechanism for running service specific functionality on client ....

[Article contains additional citation context not shown here]

A. Vahdat and T. Anderson. Transparent Result Caching. In Proceedings of the 1998 USENIX Technical Conference, New Orleans, Louisiana, June 1998.

No context found.

A. Vahdat and T. E. Anderson. Transparent Result Caching. In Proceedings of the USENIX Annual Technical Conference (USENIX '98), New Orleans, Louisiana, June 1998.

No context found.

A. Vahdat and T. E. Anderson. Transparent Result Caching. In Proceedings of the USENIX Annual Technical Conference (USENIX '98), New Orleans, Louisiana, June 1998.

No context found.

A. Vahdat and T. E. Anderson. Transparent Result Caching. In Proceedings of the USENIX Annual Technical Conference (USENIX '98), New Orleans, Louisiana, June 1998.

No context found.

A. Vahdat and T. Anderson, "Transparent result caching," in Proc. 1998.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC