Abstract: A head assembly for a magnetic tape storage device includes a head and a flexible circuit connected to the head. The flexible circuit includes a gripping portion, an end including electrical contacts for the head, conductors extending from the electrical contacts, and ink patterned onto the electrical contacts and gripping portion to form electrical paths therebetween. The ink conducts electrostatic charge from the electrical contacts and conductors to a grounded user in response to skin of the user contacting the ink.

Abstract: A method of automatically generating a lemma dictionary from a web resource may include extracting a plurality of tokens from text-based documents within the web resource, and generating a plurality of N-grams from the plurality of tokens. The method may additionally include receiving one or more filter definitions that identify valid N-grams, and filtering the plurality of N-grams using the one or more filter definitions to generate a lemma dictionary. The method may further include generating an ontology that comprises the lemma dictionary.

Abstract: Methods of providing policy based access to master keys, enabling keys to be distributed to groups of users in a secure manner while minimizing disruptions to the user in the event of changes to group membership or changes to user attributes. User attributes are identified. Policies are rewritten in terms of user attributes. New unique user attribute keys are generated for each attribute for each user. An access tree is constructed with user attribute keys as leaf nodes and Boolean algebra operations as internal nodes. Shamir polynomials are used for AND nodes, and broadcast polynomials are used for OR nodes. Master keys are accessible by traversing the access tree from the leaf nodes to the root node constructing the polynomials attached to all the nodes along the access path.

Abstract: An aspect of the present disclosure provides deep packet inspection (DPI) of network packets for keywords of a vocabulary. In one embodiment, a mapping specifying association of respective keywords to corresponding unique pattern codes is maintained, with each pattern code being shorter in length compared to the corresponding keyword and being computed based on a formula. Upon receiving a network packet, a token (containing a sequence of characters) present in the network packet is first identified and the formula then applied to the identified token to generate a token code. The token is determined to match a specific keyword when the token code equals the pattern code corresponding to the specific keyword in the mapping.

Abstract: A system for the management of rack-mounted field replaceable units (FRUs) that affords the enhanced availability and serviceability of FRUs provided by blade-based systems but in a manner that accommodates different types of FRUs (e.g., in relation to form factors, functionality, power and cooling requirements, and/or the like) installed within a rack or cabinet.

Abstract: A method and apparatus for memory paging is disclosed. A system includes a plurality of processor cores each configured to initiate requests to a memory by providing a physical address without a virtual address. A first cache subsystem is shared by each of a first subset of the plurality of processor cores. Responsive to receiving a memory access request from a processor core of the first subset, the first cache subsystem determines if a physical address of the request is in a first paged region of memory with respect to the first subset. If the physical address is in the paged region, the cache subsystem is configured to access a set of page attributes for a page corresponding to the physical address from a page attribute table responsive that is shared by each of the first subset of the plurality of processor cores.

Abstract: A chip package includes an optical integrated circuit (such as a hybrid integrated circuit) and an integrated circuit that are proximate to each other in the chip package. The integrated circuit includes electrical circuits that modulate data, communicate data, and serialize/deserialize data, and the optical integrated circuit communicates optical signals with very high bandwidth. Moreover, a front surface of the integrated circuit is electrically coupled to a top surface of an interposer, and a top surface of the integrated circuit is electrically coupled to a front surface of the optical integrated circuit. Furthermore, a bottom surface of the optical integrated circuit faces the top surface of the interposer, and the front surface of the optical integrated circuit is optically coupled to an optical-fiber receptacle, which in turn is optically coupled to an optical-fiber connector.

Abstract: A system and method for a planned migration of service connections from a first database instance to a second database instance in a clustered database. In accordance with an embodiment, a data source and connection pool enable access by a software application to a service on the first database instance. The data source is associated with a parameter specifying a delay-for-planned-draining, and a parameter specifying a planned-draining-period, which are used to rebalance connections when relocating a service. When a service is stopped on the first database instance, the connection pool receives a “service down” event from a notification service, and maintains the connections up to a time corresponding to the delay-for-planned-draining. When a “service up” event is received, the connection pool spreads migration of the connections from the first to second database instance over the planned-draining-period.

Abstract: Techniques for taking one or more nodes of a multi-node database system offline are provided. Initially, the multi-node database system uses a first data access management system (e.g., a distributed lock management system) to manage access to data stored in a database. In response to determining that one or more nodes need to be taken offline or otherwise become unavailable, the remaining nodes implement a second data access management system (e.g., an affinity-based access system) to manage access to data stored in the database. After the offline nodes are brought back online, all the nodes in the multi-node database system implement the first data access management system again. The transitions between the two data access management system may involve multiple phases that involve communicating different types of information between nodes that are taken offline and nodes that are not taken offline.

Abstract: Aspects of the present disclosure involve systems and methods for passing optimization information, which may be used to optimize input/output (“I/O”) requests, from a database management system to a file system, using various protocols of a network file system. The optimization information provided by the database management system may relate to various aspects of an I/O request that may be optimized.

Abstract: A method and apparatus are disclosed for enabling nodes in a distributed system to share one or more memory portions. A home node makes a portion of its main memory available for sharing, and one or more sharer nodes mirrors that shared portion of the home node's main memory in its own main memory. To maintain memory coherency, a memory coherence protocol is implemented. Under this protocol, a special data value is used to indicate that data in a mirrored memory location is not valid. This enables a sharer node to know when to obtain valid data from a home node. With this protocol, valid data is obtained from the home node and updates are propagated to the home node. Thus, no “dirty” data is transferred between sharer nodes. Consequently, the failure of one node will not cause the failure of another node or the failure of the entire system.

Abstract: Tooltip generation for zoom feedback is described. In response to the initiation of a zoom operation, a tooltip is generated. The tooltip shows an established potential zoom level that will be used to execute the zoom operation if no other zoom level is established or selected. Multiple potential zoom levels may be shown within the tooltip before the zoom operation is executed.

Abstract: A system for providing an interactive map displays an interactive map marker that includes an associated location on the map, an associated data value, a visual indicator, and an underlying data model. The system receives an interaction from a user that modifies the visual indicator and changes the associated data value based on the visual indicator modification. The system then changes the underlying data model based on the visual indicator modification.

Abstract: The loading or operation of a specialized class may trigger the specialization of other classes. A compiler may be configured to recognize dependency relationships between generic classes and to describe the classes in terms of the type variables of the triggering types (e.g., the types and/or type parameterizations) that trigger the specialization of classes based on the specialization of a first class. A compiler may include information, such as structural references, indicating dependency relationships between classes when generating class files. Thus, the class file may include information indicating that a class extends a class resulting from applying a specialization code generator to an argument. Loading a first class may trigger the loading of a second class described by a structural description such that a specializer (and/or class loader) may apply the structural description to generate and load the second class for the particular parameterization.

Abstract: Systems and methods used in human resource management systems. The method optimizes the assignment of permissions (e.g., ability to write to a database, ability to create a new account, etc.) to jobs. The method discovers relationships between jobs, duties and privileges by accessing an organization chart that relates a plurality of jobs, a plurality of specific duties to be performed within the purview of a given job, and a plurality of permissions for the respective duties of the job. The method then flattens the organization chart to enumerate the permissions inherited by the jobs. The method proceeds to mine the inherited permissions across the jobs to optimize the sets of permissions. The sets can be optimized (e.g., minimize number of sets, maximize coverage, etc.) and named so as to be conveniently assigned (e.g., by an HR person) to a job (e.g., in the case of a new employee).

Abstract: Embodiments of the invention provide systems and methods for scheduling and executing user-restricted processes within distributed computing systems. More specifically, certain embodiments of the present invention describe systems and methods by which runtime requests to execute user-restricted processes may be received, processed, and scheduled via a user application of an enterprise system or other distributed computing system. Requests may be received via transaction pages of user applications to schedule and execute user-restricted processes. The user application, a process scheduler, and/or other components within the distributed computing system may determine user contexts associated with the requests, and may determine whether or not the requested processes may be scheduled and executed. Such processes may include, for example, report generation processes, processes to initiate system functions, and/or processes to perform system maintenance within the distributed computing environment.

Abstract: Aspects of dynamic data compression selection are presented. In an example method, as uncompressed data chunks of a data stream are compressed, at least one performance factor affecting selection of one of multiple compression algorithms for the uncompressed data chunks of the data stream may be determined. Each of the multiple compression algorithms may facilitate a different expected compression ratio. One of the multiple compression algorithms may be selected separately for each uncompressed data chunk of the data stream based on the at least one performance factor. Each uncompressed data chunk may be compressed using the selected one of the multiple compression algorithms for the uncompressed data chunk.

Abstract: A method of quantifying resolution times through external networks includes receiving, from an external network, a transmission indicating an incident. The transmission includes a first time at which an indication of the incident was posted to the external network. The method may also include generating an incident record associated with the incident, submitting the incident record to an incident management process to resolve the incident, and receiving an indication that the incident is resolved. The method may additionally include submitting a request to the external network to post the indication that the incident is resolved, and receiving, from the external network, a second time at which the indication that the incident is resolved was posted by the external network.

Abstract: A method of handling event subscriptions and notifications may include receiving a request from a client system to subscribe to an event type, and receiving a template from the client system. The template may define data that should be returned to the client system when an event of the event type occurs. In response to the event of the event type occurring, the method may also include packaging data associated with the event according to the template received from the client system to generate an event notification. The method may further include sending the event notification to the client system.

Abstract: Techniques for automating the administration of computer systems. In one set of embodiments, information can be received specifying one or more commands and a list of target computer systems. Upon receiving this information, the one or more commands can be automatically executed in parallel against the target computer systems. In certain embodiments, executing the one or more commands in parallel can include forking a child process for each target computer system, and executing the one or more commands against that target computer system in the context of the child process. Output and error information that is collected by each child process as a result of executing the one or more commands can be aggregated and made available to a system administrator upon completion. Further, error information that is generated as a result of the automated administration process itself can be stored and made available to the system administrator for review.

Abstract: In accordance with various embodiments, systems and methods are provided which facilitate composing a telecommunication application by orchestrating application components. Embodiments of the invention enhance the use of simplified programming constructs for user activities in a telecommunications network providing an invocation manager which invokes event functions in an order identified by an orchestrator. The orchestrator holds all specified event functions each with trigger chance and priority value. The event functions can be defined in Communication Beans with annotations and may be updated and selected by an XML file.

Abstract: Aspects of the present disclosure involve systems and methods that enable the dynamic execution of code library calls using external system data. A processing device processes unstructured data to generate structured data that optionally may be filtered. The processing device uses the structured data and one or more parameter groupings to execute various commands associated with a code library of an external system.

Abstract: A scalable concurrent queue includes a central queue associated with multiple temporary queues for holding batches of nodes from multiple producers. When a producer thread or service performs an insertion operation on the scalable concurrent queue, the producer inserts one or more nodes into a batch in one of the multiple temporary queues associated with the central queue. Subsequently, the producer (or another producer) inserts the batch held in the temporary queue into the central queue. Contention between the multiple producers is reduced by providing multiple temporary queues into which the producers may insert nodes, and also by inserting nodes in the central queue in batches rather than one node at a time. The scalable concurrent queue scales to serve large number of producers with reduced contention thereby improving performance in a distributed data grid.

Abstract: A chip package includes an integrated circuit and an optical integrated circuit (such as a hybrid integrated circuit) with an optical source and/or an optical receiver. The integrated circuit and the optical integrated circuit may be proximate to each other on opposite sides of an interposer in the chip package. Moreover, the integrated circuit may include a driver circuit of electrical signals for the optical source and/or a receiver circuit of electrical signals from the optical receiver. Furthermore, the optical integrated circuit may be positioned in a hole or an etch pit in a substrate, and an alignment feature may mechanically couple the substrate to an optical-fiber assembly, so that the optical-fiber assembly is positioned relative to the interposer and the optical integrated circuit. In particular, the optical-fiber assembly may partially overlap the interposer, so that optical signals are provided and/or received from the optical integrated circuit through the interposer.

Abstract: When an unsafe port with a loss of signal is detected, a transceiver may enable one laser in a group of lasers associated with the unsafe port and may disable the remaining lasers. Then, the transceiver may instruct a transmitter associated with the one laser to transmit an optical signal on the unsafe port using a reduced transmit power that is less than a threshold value associated with the Class 1 conditions and at a different time than enabled lasers in other groups of lasers. Alternatively, for a safe port on which valid communication is received, the transceiver may enable lasers in a group of lasers associated with the safe port. Then, the transceiver may instruct transmitters associated with the lasers in this group of lasers to transmit optical signals on the safe port using a normal transmit power for the lasers that is greater than the threshold value.

Abstract: Embodiments of the present invention relate to techniques for creating policies. A plurality of objects representative of semantic objects are provided to a user. An arrangement of a subset of the objects, the arrangement representative of a policy, is received. The arrangement is converted to instructions for implementation by an application configured to implement policies. One or more of the objects may include fields and/or controls for specifying criteria of semantic objects represented by the objects.

Abstract: A distributed order orchestration system provides a plurality of representations of a product, such as a sales-centric representation of a product and a fulfillment-centric representation of a product. The distributed order orchestration system captures an order associated with the product, and creates a sales order based on the sales-centric representation of the product. The distributed order orchestration system then transform the sales order into a fulfillment order, where the fulfillment order is based on the fulfillment-centric representation of the product. As part of the transformation, product information that is associated with the sales order is transformed into product information that is associated with the fulfillment order, using one or more product transformation rules.

Abstract: A system is provided that designs a supply chain network. The system identifies a service level agreement metric definition. The system further generates a grid including cells, where the grid is located over a representation of a geographical region. The system further computes a service level agreement metric for each cell of the grid using the service level agreement metric definition. The system further selects cells that have the largest service level agreement metrics and that have not been previously selected. The system further positions a supply chain network resource at a center of the selected cells. The system further computes a service level agreement based on the service level agreement metric of the selected cells.

Abstract: Techniques for executing versioned memory access instructions. In one embodiment, a processor is configured to execute versioned store instructions of a first thread within a first mode of operation. In this embodiment, in the first mode of operation, the processor is configured to retire a versioned store instruction only after a version comparison has been performed for the versioned store instruction. In this embodiment the processor is configured to suppress retirement of instructions in the first thread that are younger than an oldest versioned store instruction until the oldest versioned store instruction has retired. In some embodiments, the processor is configured to execute versioned store instructions of a given thread within a second mode of operation, in which the processor is configured to retire outstanding versioned store instructions before a version comparison has been performed.

Abstract: Techniques are disclosed for deploying services in a server cluster environment. Certain techniques are disclosed for deploying services to a cluster based on a replication policy that includes a plurality of configurable parameters. In some embodiments, the configurable parameters (also referred to herein as replication factors) can define a number of nodes to which a service is to be deployed, a number of nodes to which a service is to be prepared, and/or a number of nodes to which a service is replicated. Based on the configurable parameters, the replication policy enables users and/or cluster providers to guarantee different levels of performance and/or reliability.

Abstract: Methods, systems, and computer-readable storage mediums are presented for aggregating multiple values for an object using database rows. Each value added to the database for the object may include a same primary key and a unique sequence key. This allows multiple values to be added for an object concurrently by simply adding additional rows to the database table. After values have been added, a process can identify multiple rows having the same primary key and aggregate the values stored in those rows together into a single database row having the primary key. This can be done as part of a garbage-collection routine or in response to other processes attempting to add additional values for the object.

Abstract: A repository server that provides stored copies of Web-accessible documents A client of the repository server may register a document in the repository server. The repository server makes a copy of the registered document and returns a repository URL for the copy to the client. The repository URL may be used to fetch the copy from the repository URL. Registration further relates the stored copy to its document URL, to an identifier for the stored copy, to a fingerprint that is a condensed representation of the stored copy's content and can be used to determine degrees of similarity other than match-no match, and to a set of stored copies having similar content. The fingerprints are used to compute similarity. The similarity computation further employs comparisons of links in the documents and of document URLS to determine whether it is necessary to use the fingerprints to compute similarity.

Abstract: Systems and methods for guaranteeing the event order for multi-stage processing in distributed systems are disclosed. In some examples, a warm-up period is used to accurately determine a starting point for ordered events of an event stream. Skip-beats may be utilized as dummy events so that the event processor does not wait too long for events that were filtered out at earlier stages.

Abstract: In some examples, a service bus may include one or more computing devices configured to translate requests from one or more client applications to one or more proxy services. According to at least one example, a system may create a resource object based at least in part on a template. The template may be associated with configuration information that may identify at least one parameter associated with a third-party application. In some examples, the third-party application may be a web service or a client application. Additionally, the system may generate a message flow based at least in part on the resource object. The system may also process a request, from the third-party application, based at least in part on the message flow. Further, the system may provide a graphical user interface (GUI) associated with the template, the GUI configured to enable input of the configuration information.

Abstract: Techniques for managing privileged accounts via a privileged access management service are provided. In some examples, the service may be configured with a plug-in framework for accessing secure resources. In some aspects, plug-in code for implementing a workflow that includes step-up validation associated with a user attempting to access at least one secure resource may be received. Access to the at least one secure resource may be provided when the user is authenticated with respect to the service. In some examples, a request to access a second secure resource may be received. Additionally, in some examples, the workflow to perform the step-up validation may be implemented at least in response to the request to access the second secure resource. The workflow implemented based at least in part on an attribute associated with the request.

Abstract: Embodiments of the present invention provide methods, system and machine-readable media for dynamically providing identity management or other services. According to one embodiment, dynamically providing services can comprise receiving a request related to an unknown principal. A service to which the principal is known can be selected. Once a service to which the principal is known has been located, an identity management result can be obtained from the selected service. The method can further comprise determining based on the identity management result whether the principal is authorized to access a requested resource. In response to determining the principal is authorized, the requested resource can be accessed.

Abstract: Methods, systems, and computer readable media for modeling packet technology services using a packet virtual network (PVN) are provided. In some aspects, and at a computing platform, a method includes providing a plurality of flow interface objects. Each flow interface object can be associated with a physical device. The method further includes configuring the flow interface objects into a PVN, wherein the flow interface objects represent network devices or device sub-interfaces. The method further includes displaying the PVN for visually modeling an Ethernet and/or a non-Ethernet packet technology service. An exemplary system includes a computing platform having at least one processor and a memory, the computing platform being configured to access a plurality of flow interface objects stored in the memory and a PVN)-modeling module (PVN-MM) for grouping the flow interface objects into a PVN for modeling a service.

Abstract: The disclosed embodiments provide a remembered set implementation for use during an incremental garbage collection, wherein the implementation includes a trie that can be dynamically coarsened to conserve memory. During operation, responsive to storing a reference into a location in a referenced memory area during the execution of a software program, the system finds, within a trie that serves as a remembered set for the referenced memory area, a particular entry that corresponds to a particular address range that covers the location. The system then marks the particular entry to indicate that the particular address range should be processed during a garbage collection. Based on a policy, the system then coarsens a particular subtree of the trie in which the particular entry is stored. Next, during the garbage collection, the system processes a particular larger address range when a root entry of the particular subtree is visited.

Abstract: Techniques for replacing a token with a mask value are disclosed. In an embodiment, a masking engine is implemented by a cloud application. The cloud application receives, from a client device, a request for a data value associated with a data field. The masking engine determines that a set of data stored in association with the data field corresponds to a token format. The cloud application transmits a mask value, rather than the set of data, towards the client device. In an embodiment, a masking engine is implemented by a client device and/or a data residency proxy. The masking engine receives a set of data for display at an interface of the client device. The masking engine determines that the set of data corresponds to a token format. The masking engine causes a mask value, rather than the set of data, to be displayed at the interface.

Abstract: Systems and methods for efficient thread arbitration in a threaded processor with dynamic resource allocation. A processor includes a resource shared by multiple threads. The resource includes an array with multiple entries, each of which may be allocated for use by any thread. Control logic detects a load miss to memory, wherein the miss is associated with a latency greater than a given threshold. The load instruction or an immediately younger instruction is selected for replay for an associated thread. A pipeline flush and replay for the associated thread begins with the selected instruction. Instructions younger than the load instruction are held at a given pipeline stage until the load instruction completes. During replay, this hold prevents resources from being allocated to the associated thread while the load instruction is being serviced.

Abstract: Techniques for circuit concurrent gate sizing and repeater insertion considering the issue of size conflicts are described herein. Certain of these techniques can be directed to coupled gates within levels of a levelized circuit falling within a coupling window defined by a minimum slack gate and adjacent gates coupled to the minimum slack gate with an adjacency parameter less than a predefined adjacency limit.

Abstract: A method, system, and computer program product for enterprise analytics. The method commences by requesting and receiving bank statement information from a bank computer where the bank statement information comprises at least a then current balance, then accessing a transaction repository to retrieve a transaction pipeline comprising deposit transactions and/or withdrawal transactions to be applied at future times to the then current balance. The method continues by forecasting a balance by applying at least a portion of the transaction pipeline to the then current balance. The resulting forecasted balance and reports are displayed for review by a user. The displays can include desktop or mobile displays or both. Transactions that involve multiple currencies can be converted into a selected currency of choice, and the balances and transactions are displayed in the denomination of the selected currency.

Abstract: An embodiment of the present disclosure receives a respective set of input messages from a corresponding source system, with the messages of each set of input messages being received in a corresponding chronological (time) order. The input messages are processed using parallel processing approaches to generate corresponding output messages, with each output message being generated for delivery to a corresponding destination system. The destination system for each output message is identified only during such processing of the input messages. The output messages are queued and an output message to a destination system is transmitted only after output messages of all prior input messages from the same source system to the destination system are transmitted, while permitting output messages generated by processing of input messages from different source systems being permitted to be transmitted out-of-sequence.

Abstract: The SIP server can be comprised of an engine tier that is designed for high throughput processing of SIP communications and a state tier that maintains state information for access by the engine tier. A Java-based offline charging application can be deployed on the SIP server engine tier in order to allow the SIP server to generate offline charging events and transmit them to a charging data function (CDF) server. The state data for the offline accounting session can be maintained on the state tier. Furthermore, timers set to accounting interim interval value can be used to indicate expiration times for the accounting session. The SIP server can use the timers to send periodic interim requests in order to update and maintain the accounting session open.

Abstract: A system and method for providing collaboration communities in a computer portal environment. The system comprises a community portal tool for creating community portals, portal applications, or portal web applications. In one embodiment, the portal can be one of a number of enterprise applications on server software. The portal can include a community context that can have an indication of members of the community. The indication can be, for example, a list of the members or pointer to a list of members. The community portal can be constructed from portal elements and the community context. The community portal tool can also be used to select a community portal template from a portal component library. The template can include standard elements from the construction of a community portal. A standardized look and feel, layout, and components such as a calendars can be in a community template.

Abstract: Tools are provided to assist users of cloud computing systems to create, monitor, and debug the resource instances they need. Specifically, resource instances are created in a cloud computing system based on a blueprint that is provided to a blueprint processor. The blueprint declaratively defines the resources that the user needs to have created in the cloud. Based on the information contained in blueprint, the blueprint processor causes the cloud to orchestrate the creation of resource instances, doing so in parallel when possible, and to configure those resource instances as specified in blueprint. Techniques are described for generating a timeline to convey status during the deployment process, and for performing debugging operations.

Abstract: Techniques for managing accounts are provided. An access management system may check out credentials for accessing target systems. For example a user may receive a password for a period of time or until checked back in. Access to the target system may be logged during this time. Upon the password being checked in, a security account may modify the password so that the user may not log back in without checking out a new password. Additionally, in some examples, password policies for the security account may be managed. As such, when a password policy changes, the security account password may be dynamically updated. Additionally, in some examples, hierarchical viewing perspectives may be determined and/or selected for visualizing one or more managed accounts. Further, accounts may be organized into groups based on roles, and grants for the accounts may be dynamically updated as changes occur or new accounts are managed.

Abstract: A method for transport layer lookup involves receiving a first incoming transport layer packet, and searching a pointer cache for a first matching transport layer data structure including state information corresponding to the first incoming packet. The pointer cache includes pointer cache lines, each of which stores at least one pointer to a subset of global transport layer data structures. The method further involves returning the state information corresponding to the first incoming packet using the first matching transport layer data structure when a pointer cache hit occurs, receiving a second incoming transport layer packet, searching the pointer cache for a second matching transport layer data structure including state information corresponding to the second incoming packet, and searching the plurality of global transport layer data structures in main memory to obtain the matching second transport layer data structure, when a pointer cache miss occurs.

Abstract: Techniques for managing network-connected objects are provided. In some examples, code for accessing a network-connected object may be received. The code may be configured to enable generation of an application programming interface method. In some aspects, account information associated with a user may be stored. A particular method call corresponding to the application programming interface method may be received from a computer device of the user. The particular method call may include a request to access the network-connected object. In some examples, the request to access the network-connected object may be authenticated based at least in part on the account information. Additionally, in some examples, an instruction to the network-connected object may be provided over a network if the request is authenticated.

Abstract: A database server exploits the power of compression and a form of storing relational data referred to as column-major format, to store XML documents in shredded form. The column values that are to be stored for shredded XML documents are separately analyzed for a XML document to determine whether to store a particular column in column-major format or row-major format, and what compression technique to use, if any.