International Business Machines Corporation

Rashmi Narasimhan

Nikhil Hegde

Ajay Kumar Mahajan

Anis M Abdul

Kavitha Vittal Murthy Baratakke

Sivakumar Krishnasamy

Nicholas E Bofferding

Kavitha V Baratakke

Vasu Vallabhaneni

Dwip N Banerjee

Jos Manuel Accapadi

Adekunle Bello

Aruna Yedavilli

Jeffrey Paul Messing

Michael John Jones

Patrick Tam Vo

Radhika Chirra

Atit D Patel

David Zhiwei Dai

Nicholas Anthony Pietraniec

Kavitha V M Baratakke

Alexander Medvedev

Shouchun Li

Richard Jamie Knight

In a method for regulating network bandwidth in a virtualized computer environment, a computer having a hypervisor program receives a request from a first virtual client to transmit data. In response, the computer transfers the data from a memory of the first virtual client to a memory of a virtual server. The computer receives an error notification from a shared virtual network adapter of the virtual server, indicative of insufficient network bandwidth available to transmit the data. In response, the computer notifies the first virtual client that insufficient network bandwidth is available to transmit the data.

Regulating network bandwidth in a virtualized environment

Provided is reporting session information for a session upstream from a server. A data processing system receives a session handshake from a first client for a session between the server and the first client. The system receives a session-descriptor for a session between a second client and the first client before handling a nested session. The system determines whether a user has made an upstream session request. The system obtains session data from an upstream data structure, responsive to a determination that the user has made an upstream session request. The system renders the session data based on the upstream session request.

Remote session management

A mechanism, in a data processing system, is provided for logical partition defragmentation. The mechanism gathers resource requirements for a plurality of logical partitions running in a plurality of power domains within one or more servers. The mechanism determines optimal hardware utilization for the plurality of logical partitions. The mechanism migrates one or more of the plurality of logical partitions to run in a subset of the plurality of power domains such that at least one power domain within the plurality of power domains is unused. The mechanism puts the at least one unused power domain in a low power state.

Logical partition defragmentation within a data processing system

A method, system, and computer program product for a self-adjusting application heartbeat management process are provided in the illustrative embodiments. A first application receives from a second application, a response message in response to a heartbeat message. Information is identified in the response message usable for computing a delay in a future response message for a future heartbeat message, such that the information communicates to the first application a system condition existing in the second data processing system where the second application is executing. A response time is adjusted using the information to form a modified response time. The future response message is received after the response time.

Self-adjusting application heartbeat management

An approach is provided to improve network efficiency. A send segment size, such as a maximum segment size (MSS), that corresponds to data segments being sent to a receiver over a computer network. A data block is identified in a send buffer and the data block includes more than one data segments. Based on the determined send segment size, all but a remaining data segment of the data segments are sent to the receiver. The sent data segments are each the determined send segment size and the remaining data segment is smaller than the send segment size. The remaining data segment is sent to the receiver in response to identifying that the remaining data segment is a portion of the data block.

Smart nagling in a TCP connection

A kernel extension is configured to intercept a call to associate a socket with a port of a node in a network. The call originates from a kernel of the node. The kernel extension is configured to determine the port from the call. The kernel extension is configured to determine that the port is one of a plurality of ports for which the node has authority to modify firewall rules of a firewall of the network. The kernel extension is configured to modify firewall rules maintained by the firewall to allow communications for the port to the node through the firewall.

Secure firewall rule formulation

A mechanism is provided to enhance the scalability of network caching capabilities. All network client applications running on a partition in a virtualized environment are configured to query information from a single virtual input/output server (VIOS) network caching daemon. Thus, the illustrative embodiments provide a 1:n model where a VIOS partition has a network caching daemon, and each of the n partitions uses the network caching daemon of the VIOS partition. The mechanism of the illustrative embodiments only requires the system administrator to control one copy of the local files on the VIOS server. The system administrator does not need to monitor all of these local files on each individual partition. The system administrator consolidates the entries used by the different individual partitions into one single file on the VIOS server side.

Enhancing the scalability of network caching capability in virtualized environment

Communication between nodes on distinct physical computer systems typically involves fragmentation of packets, such that packet size complies with a communication protocol's specified maximum transmission unit (MTU). However, communications between logical partitions on the same computer system ('virtual nodes') do not have to comply with a MTU. A computer system can handle packet sizes for intra-system communications between virtual nodes larger than the MTU. Functionality can be implemented to determine that a packet's source and destination are virtual nodes on the same physical machine. This can enable the system to virtually transfer packets for intra-system communication without fragmentation.

Discriminatory MTU fragmentation in a logical partition

A computer implemented method, a tangible computer medium, and a data processing system are provided for waking a blade server from an operational state of reduced power. When server blade enters the state of reduced power, a service firmware configures a multi-port blade switch of the server blade to direct incoming packets to the service firmware. The service firmware then polls for receipt of a Wake-on-Lan magic packet. When the Wake-on-Lan magic packet is received by the service firmware, the service firmware reconfigures the multi-port blade switch to direct incoming packets to a network interface card of the server blade. The service firmware then initiates a reboot of the server blade.

Wake on Lan for blade server

A method, system and computer program product for remotely debugging a malfunctioning node controller of a node in a distributed node network through a functioning node controller of the same node. The method comprises establishing a serial link between the malfunctioning node controller and a functioning node controller and configuring the functioning node controller as a virtual console by the remotely-located central data processing system (DPS). The method further includes receiving, via an internal Fru Support Interface (FSI) link, serial data from the malfunctioning node controller through the virtual console, and debugging, by the DPS, a failure condition of the malfunctioning node controller, in response to receipt of the serial data through the virtual console.

Method and system for remote node debugging using an embedded node controller

A computer implemented method, a tangible computer readable medium, and a data processing system intelligently propagate link status information received by a blade server to the various ports of an embedded multi-port switch. The link status of a switch port in an external switch module can be communicated to the operating systems of individual blade servers that are affected by that link status. When an external switch module is unplugged from a server blade chassis, the bus controller broadcasts a link down event, such as a link down interrupt, to the individual server blades where it is received by the embedded multi-port switch for those server blades. The embedded multi-port switch translates the link down interrupt into a hardware link down event, and forwards the hardware link down event to the other elements connected to the embedded multi-port switch.

Interrupt-driven link status feedback mechanism for embedded switches

A method for enabling a Node Controller (NC), which claims a duplicate or invalid service processor Node Controller Identification (NCID) in a distributed service processor system, to be integrated into the system includes reading an NCID by the NC after the NC is booted, saving the NCID into a non-volatile storage and broadcasting an NC Present Message (NPM) to a System Controller (SC) repeatedly until the SC initiates communication, updating the NCID for the NC in the non-volatile storage when the NC receives an NCID change message from the SC and rating any future NPM as a new NCID, and checking a record of a new NC when the SC receives the NPM from the NC. If the SC has a record of a recorded NC with the same NCID as the new NC, then the SC checks its role as a primary SC. If the SC does not have the record of the recorded NC with the same NCID as the new NC, then the SC checks validity of the NCID.

Method of handling duplicate or invalid node controller IDs in a distributed service processor environment

A memory registration abstraction layer includes decision logic for dynamic, efficient determination of memory copy versus registration. The exemplary embodiments apply to direct memory access environments. The memory registration abstraction layer also includes tracking logic for monitoring the usage history for certain ranges of memory registration requests. The registration logic chooses to copy memory or map memory based on training and observation of application behavior, taking the more efficient performance path without requiring changes to the application.

Growing community of inventors

Round Rock, TX, United States of America

Rashmi Narasimhan