International Business Machines Corporation

Rohit Dilip Kelapure

Erik John Burckart

Snehal S Antani

Dana Lea Price

Andrew J Ivory

Aaron Kyle Shook

Lisa M W Bradley

James William Stopyro

Matthew J Sykes

Junjie Cai

Kevin Grigorenko

Todd Eric Kaplinger

Hariharan N Venkitachalam

Wim De Pauw

Paul Komar

Maneesh Mehra

Madhu K Chetuparambil

Kevin Edward Vaughan

Maxim Avery Moldenhauer

Andrew Henry Richter

Erinn Elizabeth Koonce

Aravind Srinivasan

Jai Singh Arun

David N Brauneis, Jr

Dana Rieger Thalheimer

Paul Anderson

Gautam Singh

Christian Steege

Filip R Zawadiak

Jeffrey A Lee

Ching C A Chow

Ching Chi Andrew Chow

Ching Chi Chow

Amith N Kashyap

Embodiments of the present invention provide a method, system and computer program product for the dynamic structural management of an n-Tier distributed caching infrastructure. In an embodiment of the invention, a method of dynamic structural management of an n-Tier distributed caching infrastructure includes establishing a communicative connection to a plurality of cache servers arranged in respective tier nodes in an n-Tier cache, collecting performance metrics for each of the cache servers in the respective tier nodes of the n-Tier cache, identifying a characteristic of a specific cache resource in a corresponding one of the tier nodes of the n-Tier crossing a threshold, and dynamically structuring a set of cache resources including the specific cache resource to account for the identified characteristic.

Dynamic structural management of a distributed caching infrastructure

Embodiments of the present invention provide a method, system and computer program product for enforcing performance goals in an n-Tier distributed caching infrastructure. In an embodiment of the invention, a method of enforcing performance goals in an n-Tier distributed caching infrastructure can include establishing a communicative connection to multiple different cache servers arranged in respective tier nodes in an n-Tier cache. The method also can include collecting performance metrics for each of the cache servers in the respective tier nodes of the n-Tier cache and identifying a cache server amongst the cache servers in respective tier nodes of the n-Tier demonstrating a performance likely to breach at least one term of an SLA. Finally, the method can include applying a remedial measure to the identified cache server.

Defining enforcing and governing performance goals of a distributed caching infrastructure

A system and method that performs retrofitting and autowiring of service bindings by generating a plurality of retrofitted references from a plurality of resource references and autowiring the plurality of retrofitted references into an application. The system and method deploys the application to an environment in response to the autowiring of the plurality of retrofitted references.

Autowiring location agnostic services into application software

A code base, including a code base history, is searched for code term(s) and change(s) that have been made to the code term(s). The code base includes, but is not limited to, multiple: (i) instances of code; (ii) product releases; (iii) code components; and/or (iv) code drafting projects.

Analysis of source code changes

Resolving virtual machine (VM) issues, by executing VM and operating system (OS) diagnostic monitors, including, monitoring a set of VM and OS health status metrics of a system at a first level, analyzing data of the monitored health status metrics to determine that an instability has occurred when the data exceeds defined bounds for the health status metrics, responding to the instability by monitoring additional VM and OS health status metrics, whereby a level of monitoring of the system is increased from the first level to a second level, greater than the first level, identifying the instability, repairing the system by taking corrective action based on the identified instability; and removing at least one of the set of monitoring and profiling tools to reduce the level of monitoring to a third level once the instability has been resolved, wherein the third level is less than the second level.

Monitoring and resolving deadlocks, contention, runaway CPU and other virtual machine production issues

A method, system and computer program product for providing high availability to a hybrid application server environment containing non-Java&#xae; containers. Each hybrid application server in the cluster includes a Java&#xae; container and a non-Java&#xae; container hosting Java&#xae; and non-Java&#xae; applications, respectively. Upon detecting the non-Java&#xae; container becoming unavailable (failing), an object, such as an MBean, identifies and deactivates those Java&#xae; application(s) that are dependent on the non-Java&#xae; application(s) deployed in the unavailable non-Java&#xae; container using dependency information stored in an application framework. The deactivated Java&#xae; application(s) are marked as being unavailable. A routing agent continues to send requests to those Java&#xae; application(s) that are not marked as being unavailable within that hybrid application server containing the unavailable non-Java&#xae; container. As a result of not deactivating the entire hybrid application server containing the unavailable non-Java&#xae; container, unimpacted applications continue to service requests thereby optimally using the resources.

Providing high availability to a hybrid application server environment containing non-java containers

Embodiments of the present invention provide a method, system and computer program product for dual timer fragment caching. In an embodiment of the invention, a dual timer fragment caching method can include establishing both a soft timeout and also a hard timeout for each fragment in a fragment cache. The method further can include managing the fragment cache by evicting fragments in the fragment cache subsequent to a lapsing of a corresponding hard timeout. The management of the fragment cache also can include responding to multiple requests by multiple requestors for a stale fragment in the fragment cache with a lapsed corresponding soft timeout by returning the stale fragment from the fragment cache to some of the requestors, by retrieving and returning a new form of the stale fragment to others of the requestors, and by replacing the stale fragment in the fragment cache with the new form of the stale fragment with a reset soft timeout and hard timeout.

Dual timeout caching

Embodiments of the present invention address deficiencies of the art in respect to workload management of backend systems and provide a novel and non-obvious method, system and computer program product for limiting a request rate from a client based on a health state of a server system. In one embodiment of the invention, a computer-implemented method for limiting a request rate from a client based on a health state of a server system that includes monitoring a server system, determining a measure of health of the server system, sending the determined measure of health to a client submitting a request to the server system and receiving a subsequent request from the client that is delayed by a time period, wherein the time period is based at least in part upon the determined measure of health. The computer-implemented method can further include receiving server system CPU utilization, determining an approximate predicted real server response time of a transaction and sending the determined approximate predicted real server response time of a transaction to a client submitting a request to the server system.

Reducing workload on a backend system using client side request throttling

A system and method for searching a bytestream or other string in a case insensitive manner. In a preferred embodiment, the present invention includes an ASCII tree that associated both upper and lower case letters of an incoming header key (or other symbols) with nodes. When enough nodes are gathered, a target word is found.

Search via fast case insensitive ASCII tree

This Sampling Object Cache System ('SOCS') estimates the size of an in-memory heap-based object cache without the need to serialize every object within the cache. SOCS samples objects at a user-determined rate and then computes a 'sample size average' for each type of class&#x2014;whether a top class, type of top class or non top class. Using these sample size averages, a statistically accurate measure of the overall size of the cache is calculated by adding together the total size of the objects in the cache for each class type.

Estimating the size of an in-memory cache

Growing community of inventors

Durham, NC, United States of America

Rohit Dilip Kelapure