Emc Corporation

Emc IP Holding Company LLC

Kenneth Fickie

Amanuel Ronen Artzi

Sesh Jalagam

Hanna Yehuda

Stephen Glen Graham

Rajesh Nandyalam

James Edward Lavallee

Alexander Dubrovsky

Ethan D Roberts

Steven Teng

Richard T Simon

Pramod Kulyadi Pai

Zakir G Magdum

Rony R Baenziger

Saravanan Chithambaram

Harrison Roberts

Steven Tanaka

Kent Bair

Shankar Jagannathan

Denis Kennelly

An improved technique involves using a systems dynamics model in an information technology (IT) data center in order to determine an optimal distribution of data among data center devices. Along these lines, a data center control server takes measurements of devices across an IT data center over time and compares these measurements to quantities specified in set points (e.g., a service level agreement (SLA)) to produce deviations. The data center control server then inputs the deviations from the set points into a systems dynamics engine that determines a configuration of the devices in the IT data center so that output from the IT data center satisfies a set of constraints, including those specified in the SLA. The data center control server then configures the IT data center devices according to the configuration to send incoming data along the specified data paths.

System modeling of data centers to drive cross domain characterization, automation, and predictive analytics

Modeling a data path of a network environment is disclosed. Components are identified at a source location of a data path, at a target location of the data path, and connections there between. The network environment includes a plurality of domains, and the data path traverses the plurality of domains. Components and the connections there between are assigned to one of: an input-output path pattern, a virtualizer pattern, and a meta pattern. The patterns represent common elements of components of the network environment. The patterns and relationships between patterns are defined to repeat within and across domains in the network environment. A model of the data path is created from the assigned patterns and the defined relationships between patterns. The created model of the data path is then presented.

Generalized virtualizer IO path model and general virtual transformation model for storage assets

A method, and computer program product for computing a change plan are presented. A model of a current configuration is identified, the model including modeled system assets and modeled tasks. A request to change the current configuration to a new configuration is received. The request to change the current configuration to generate a plurality of valid strategies to be used in a genetic programming parse tree corresponding to modeled assets and modeled tasks of said model is applied to the model. The request to change the current configuration is applied to at least one of said valid strategies to generate a plurality of potential change plans. The change plans are evolved according to genetic programming principles to a preferred change plan.

Method and apparatus for computing a change plan using genetic programming and parse tree

Configurations herein present a multi-segment perspective of the SAN that models the SAN as an interconnection of protocol independent segments, each operable according to a particular transport mechanism and connected to other segments by an intersegment device such as a network bridge. Each segment is operable to include a fabric of manageable entities, such as storage arrays, switches, and hosts, that operate according to particular protocols and transport mechanisms employed by the fabric. Groups of manageable entities operable according to a particular protocol or transport mechanism are enumerated as a separate segment, and managed as a separate entity from the management application. The intersegment connections occur via an inter segment device operable to perform the mapping or translation from the respective address space of the segments. Therefore, a path from an initiator to a receiver manageable entity may be enumerated and managed by the management application as a multi-segment connection.

Methods and apparatus for modeling a storage area network

Methods and apparatus for removing data from a cache are disclosed. The data may be one or more objects that are stored in a first cache. Each object stored in the first cache has an associated reference that is stored in a second cache. Each such reference has an associated timer. As a program accesses an object stored in the first cache, the timer associated with its reference is reset. A harvester process parses through the second cache over a period of time, locking a section of the second cache and determining if any timers associated with references therein have timed out. The harvester process removes any such timed-out timers from the second cache. A garbage collection process parses through the first cache, removing any objects therein that do not have an associated reference in the second cache.

Methods and apparatus for removing data from a cache

A processing device generates a relationship view of managed entities in a storage area network based on objects stored in a relational database. Each managed entity (e.g., file system, database, volume, storage device, etc.) in the storage area network has a corresponding managed object stored in the relational database. Generally, managed objects of a particular type are categorized into one or more tables. Such tables are related to one another by foreign key attributes, which represent relationships between types of managed objects (and thus a relationship between corresponding managed entities). Relationships among the managed objects in the relational database may be generally classified as either association or containment. Association identifies a logical relationship between managed objects. Containment identifies groups of managed objects that are logically associated with each other. The processing device uses these identified relationships to generate the appropriate relationship view for a selected managed entity.

Methods and apparatus for managing network resources via use of a relationship view

Growing community of inventors

Ashland, MA, United States of America

Kenneth Fickie