公开(公告)号：US09426076B2

公开(公告)日：2016-08-23

申请号：US13610376

申请日：2012-09-11

申请人： Seraphin B. Calo ,  William Cornejo ,  Thai F. Le ,  Erich M. Nahum ,  Maroun Touma ,  Dinesh C. Verma

发明人： Seraphin B. Calo ,  William Cornejo ,  Thai F. Le ,  Erich M. Nahum ,  Maroun Touma ,  Dinesh C. Verma

IPC分类号： G06F15/177 ,  H04L12/801 ,  H04L12/803

CPC分类号： H04L41/0816 ,  H04L43/04 ,  H04L43/0841 ,  H04L43/0888 ,  H04L47/10 ,  H04L47/12 ,  H04L47/122 ,  H04L47/127 ,  H04L61/2015 ,  H04L61/6022 ,  H04L67/42

摘要： Dynamic middleboxes include a network monitor module configured to determine a quality-affective factor in a connection between a client and a server in a network; a processor configured to compare the quality-affective factor to a threshold to determine whether the connection would benefit from a network processing function; and a network control module configured to reconfigure the network to redirect the connection to a middlebox that performs the network processing function if the connection would benefit and further configured to reconfigure the network to exclude the middlebox from the connection if the connection would not benefit.

摘要翻译： 动态中间箱包括网络监视器模块，其被配置为确定网络中客户端和服务器之间的连接中的质量因素; 处理器，被配置为将所述质量因素与阈值进行比较，以确定所述连接是否将受益于网络处理功能; 以及网络控制模块，被配置为重新配置所述网络以将所述连接重定向到执行所述网络处理功能的中间件，如果所述连接将受益并且进一步被配置为如果所述连接不受益，则重新配置所述网络以从所述连接中排除所述中间件。

公开(公告)号：US08831026B2

公开(公告)日：2014-09-09

申请号：US10804516

申请日：2004-03-19

申请人： Arun Kwangil Iyengar ,  Erich M. Nahum ,  Bianca Schroeder

发明人： Arun Kwangil Iyengar ,  Erich M. Nahum ,  Bianca Schroeder

IPC分类号： H04L12/54 ,  H04L29/08

CPC分类号： H04L67/322 ,  H04L67/2809 ,  H04L67/2814

摘要： Techniques are provided for dynamically scheduling requests in data processing systems in accordance with differentiated service levels. In a first aspect of the invention, a technique for processing a request to at least one server comprises the following steps. A request is received. Then, submission of the request to the at least one server is scheduled based on: (i) a quality-of-service (QoS) class assigned to a client from which the request originated; (ii) a response target associated with the QoS class; and (iii) an estimated response time associated with the at least one server. The technique may further comprise the step of withholding the request from submission to the at least one server when the request originated from a client assigned to a first QoS class to allow a request that originated from a client assigned to a second QoS class to meet a response target associated therewith.

摘要翻译： 提供技术以根据差异化的服务水平在数据处理系统中动态调度请求。 在本发明的第一方面中，一种用于处理对至少一个服务器的请求的技术包括以下步骤。 收到请求。 然后，基于以下方式调度对所述至少一个服务器的请求的提交：（i）分配给发起请求的客户端的服务质量（QoS）类; （ii）与QoS类相关联的响应目标; 和（iii）与所述至少一个服务器相关联的估计响应时间。 该技术可以进一步包括当请求从分配给第一QoS类的客户端发起以允许从分配给第二QoS类别的客户端发起的请求满足一个请求时将该请求提交给至少一个服务器的步骤 与之相关的反应目标。

公开(公告)号：US07493427B2

公开(公告)日：2009-02-17

申请号：US10891196

申请日：2004-07-14

申请人： Douglas Morgan Freimuth ,  Elbert C. Hu ,  Ronald Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John Michael Tracey

发明人： Douglas Morgan Freimuth ,  Elbert C. Hu ,  Ronald Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John Michael Tracey

IPC分类号： G06F3/00 ,  G06F15/167 ,  G06F13/00

CPC分类号： H04L49/9052 ,  G06F12/0862 ,  H04L47/621 ,  H04L49/90 ,  H04L49/9063

摘要： A number of improvements in network adapters that offload protocol processing from the host processor are provided. Specifically, an improved mechanism for handling receipt of data packets in a system utilizing an offload network adapter. The offload network adapter may include logic that permits the offload network adapter to delay notification of data reception to the host system in different ways. The advantage of delaying the notice of data packet reception to the host system is the potential for aggregation of several data packets, which can arrive immediately after the first one, for example, in a single notification. Given a stream with continuous data packet arrival, a value may be set, either statically or dynamically, for notification delay and this value may be configurable for the host system per communication socket.

摘要翻译： 提供了从主机处理器卸载协议处理的网络适配器的许多改进。 具体地说，一种用于处理利用卸载网络适配器的系统中数据分组的接收的改进机制。 卸载网络适配器可以包括允许卸载网络适配器以不同的方式将对数据接收的通知延迟到主机系统的逻辑。 将数据分组接收通知延迟到主机系统的优点是几个数据分组的聚合的可能性，其可以在第一个数据分组之后立即到达，例如在单个通知中。 给定具有连续数据分组到达的流，可以静态或动态地设置用于通知延迟的值，并且该值可以针对每个通信套接字的主机系统来配置。

公开(公告)号：US20080031258A1

公开(公告)日：2008-02-07

申请号：US11497948

申请日：2006-08-01

申请人： Arup Acharya ,  Erich M. Nahum ,  John Michael Tracey ,  Xiping Wang ,  Charles P. Wright ,  Zhen Xiao

发明人： Arup Acharya ,  Erich M. Nahum ,  John Michael Tracey ,  Xiping Wang ,  Charles P. Wright ,  Zhen Xiao

IPC分类号： H04L12/56

CPC分类号： H04L47/10 ,  H04L47/19 ,  H04L47/2441 ,  H04L47/32 ,  H04L47/801 ,  H04L47/821 ,  H04L49/90 ,  H04L65/1046 ,  H04L69/22

摘要： A method for operating a server having a maximum capacity for servicing requests comprises the following steps: receiving a plurality of requests; classifying each request according to a value; determining a priority for handling the request according to the value, such that requests with higher values are assigned higher priorities; placing each request in one of multiple queues according to its priority value; and dropping the requests with the lowest priority when the plurality of requests are received at a rate that exceeds the maximum capacity. The server operates according to a session initiation protocol.Classifying each request comprises running a classification algorithm. The classification algorithm comprising steps of: receiving a rule set, each rule comprising headers and conditions; creating a condition table by taking a union of all conditions in the rules; creating a header table by extracting a common set of headers from the condition table; extracting the relevant headers from the header table; determining a matching rule; creating a bit vector table; selecting the matching rule according to data in the bit vector table; and applying the rule to place the message in the appropriate queue.

摘要翻译： 一种用于操作具有用于服务请求的最大容量的服务器的方法包括以下步骤：接收多个请求; 根据值对每个请求进行分类; 确定根据该值处理请求的优先级，使得具有较高值的​​请求被分配较高的优先级; 根据其优先级值将每个请求放置在多个队列之一中; 并且当以超过最大容量的速率接收到多个请求时，丢弃具有最低优先级的请求。 服务器根据会话启动协议进行操作。 分类每个请求包括运行分类算法。 分类算法包括以下步骤：接收规则集，每个规则包括头部和条件; 通过规则中的所有条件联合创建条件表; 通过从条件表中提取一组公共标题来创建头表; 从头表中提取相关头文件; 确定匹配规则; 创建一个位向量表; 根据位向量表中的数据选择匹配规则; 并应用规则将消息放置在适当的队列中。

公开(公告)号：US20140040451A1

公开(公告)日：2014-02-06

申请号：US13562603

申请日：2012-07-31

申请人： Dakshi Agrawal ,  Thai V. Le ,  Erich M. Nahum ,  Vasileios Pappas

发明人： Dakshi Agrawal ,  Thai V. Le ,  Erich M. Nahum ,  Vasileios Pappas

IPC分类号： G06F15/173

CPC分类号： H04L41/12 ,  H04L43/0841 ,  H04L47/193 ,  H04L47/40 ,  H04L67/28 ,  H04L67/2809 ,  H04L67/2814 ,  H04L69/16

摘要： Methods for inserting a middlebox into a network connection include monitoring network state information in a connection between a client and a server. When the connection is idle, a connection entry is created for each device and is initialized using state information gathered by monitoring the network connection. Redirection of the network connection is activated between the client and the server such that the middlebox mediates the connection. Methods for removing a middlebox from a network connection include determining a degree of mismatch between a sequence number in a first connection between the middlebox and a client and a sequence number in a second connection between the middlebox and a server, delaying acknowledgment signals from the middlebox on a connection to decrease the degree of mismatch, and establishing a direct connection between the client and the server without mediation by the middlebox when the degree of mismatch is zero.

摘要翻译： 将中间箱插入网络连接的方法包括在客户机和服务器之间的连接中监视网络状态信息。 当连接空闲时，为每个设备创建连接条目，并使用通过监视网络连接收集的状态信息进行初始化。 在客户机和服务器之间激活网络连接的重定向，使得中间件中介连接。 从网络连接去除中间箱的方法包括确定中间箱和客户端之间的第一连接中的序列号与中间件与服务器之间的第二连接中的序列号之间的不匹配程度，延迟来自中间件的确认信号 在连接上减少不匹配的程度，并且当不匹配程度为零时，通过中间件不建立客户端和服务器之间的直接连接。

公开(公告)号：US20090222564A1

公开(公告)日：2009-09-03

申请号：US12464741

申请日：2009-05-12

申请人： Douglas Morgan Freimuth ,  Elbert C. Hu ,  Ronald Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John Michael Tracey

发明人： Douglas Morgan Freimuth ,  Elbert C. Hu ,  Ronald Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John Michael Tracey

IPC分类号： G06F15/16

CPC分类号： H04L69/166 ,  H04L49/90 ,  H04L49/901 ,  H04L69/161 ,  H04L69/163

摘要： A number of improvements in network adapters that offload protocol processing from the host processor are provided. Specifically, a mechanism for improving connection establishment in a system utilizing an offload network adapter is provided. The connection establishment mechanism provides the ability to offload connection establishment and maintenance of connection state information to the offload network adapter. As a result of this offloading of connection establishment and state information maintenance, the number of communications needed between the host system and the offload network adapter may be reduced. In addition, offloading of these functions to the offload network adapter permits bulk notification of established connections and state information to the host system rather than piecemeal notifications as is present in known computing systems.

摘要翻译： 提供了从主机处理器卸载协议处理的网络适配器的许多改进。 具体地，提供了一种用于改善利用卸载网络适配器的系统中的连接建立的机制。 连接建立机制提供将连接状态信息的连接建立和维护卸载到卸载网络适配器的能力。 由于连接建立和状态信息维护的这种卸载的结果，可以减少主机系统和卸载网络适配器之间所需的通信数量。 另外，将这些功能卸载到卸载网络适配器允许将已建立的连接和状态信息批量通知给主机系统，而不是按照已知的计算系统中存在的零碎的通知。

公开(公告)号：US06968358B2

公开(公告)日：2005-11-22

申请号：US10205024

申请日：2002-07-25

申请人： Douglas Freimuth ,  Roland Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John Michael Tracey

发明人： Douglas Freimuth ,  Roland Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John Michael Tracey

IPC分类号： G06F9/46 ,  G06F12/00 ,  G06F15/16 ,  G06F15/167

CPC分类号： G06F9/544

摘要： A method, apparatus, and computer instructions for transferring data objects. A request for a data object is received at a communication adapter. A determination is made as to whether the data object is stored in a memory in the communication adapter in response to receiving the request. The data object is transferred from the memory to a receiver of the data object through the communication adapter in which transfer of the data across a bus connecting the communication adapter to the data processing system is unnecessary in response to the data object being present in the memory.

摘要翻译： 一种用于传送数据对象的方法，装置和计算机指令。 在通信适配器处接收到对数据对象的请求。 响应于接收到请求，确定数据对象是否存储在通信适配器中的存储器中。 数据对象通过通信适配器从存储器传送到数据对象的接收器，其中响应于存储在存储器中的数据对象，不需要通过连接通信适配器到数据处理系统的总线的数据传输 。

公开(公告)号：US09231881B2

公开(公告)日：2016-01-05

申请号：US13568941

申请日：2012-08-07

申请人： Dakshi Agrawal ,  Thai V. Le ,  Erich M. Nahum ,  Vasileios Pappas

发明人： Dakshi Agrawal ,  Thai V. Le ,  Erich M. Nahum ,  Vasileios Pappas

IPC分类号： H04L12/905 ,  H04L12/893 ,  H04L29/08 ,  H04L29/06 ,  H04L12/801

CPC分类号： H04L41/12 ,  H04L43/0841 ,  H04L47/193 ,  H04L47/40 ,  H04L67/28 ,  H04L67/2809 ,  H04L67/2814 ,  H04L69/16

摘要： Middlebox systems that can enter a connection include a monitoring module to monitor information in a connection between a client and a server, a processor to determine that the connection is idle, a table configured to create a first entry for the client and a second entry for the server, where the entries are initialized using information gathered by the monitoring module, and a control module to redirect the connection between the client and server to the middlebox. Middlebox systems that can exit a connection include a processor to determine a mismatch between sequence numbers in a first connection to a client device and in a second connection to a server device. A network control module delays acknowledgment signals from the middlebox on a connection to decrease the degree of mismatch and establishes a direct connection between the client device and the server device when mismatch is zero.

摘要翻译： 可以进入连接的中型箱系统包括：监视模块，用于监视客户端与服务器之间的连接中的信息;处理器，用于确定连接空闲;配置为为客户端创建第一条目的表;以及第二条目 服务器，其中使用监视模块收集的信息初始化条目，以及控制模块，用于将客户端和服务器之间的连接重定向到中间框。 可以退出连接的中型箱系统包括处理器，用于确定与客户端设备的第一连接中的序列号之间的不匹配，以及与服务器设备的第二连接。 网络控制模块将来自连接中的中间盒的确认信号延迟以减少不匹配的程度，并且当不匹配为零时在客户端设备和服务器设备之间建立直接连接。

公开(公告)号：US08863144B2

公开(公告)日：2014-10-14

申请号：US12723914

申请日：2010-03-15

申请人： Arun Kwangil Iyengar ,  Hongbo Jiang ,  Erich M. Nahum ,  Wolfgang Segmuller ,  Asser Nasreldin Tantawi ,  Charles P. Wright

发明人： Arun Kwangil Iyengar ,  Hongbo Jiang ,  Erich M. Nahum ,  Wolfgang Segmuller ,  Asser Nasreldin Tantawi ,  Charles P. Wright

IPC分类号： G06F9/46 ,  G06F15/173 ,  G06F9/50

CPC分类号： G06F9/46 ,  G06F9/505 ,  G06F2209/501 ,  G06F2209/5019 ,  H04L67/1008

摘要： In a computer system comprising a plurality of computing devices wherein the plurality of computing devices processes a plurality of tasks and each task has a task type, a method for determining overheads associated with task types comprises the following steps. Overheads are estimated for a plurality of task types. One of the plurality of computing devices is selected to execute one of the plurality of tasks, wherein the selection comprises estimating load on at least a portion of the plurality of computing devices from tasks assigned to at least a portion of the plurality of computing devices and the estimates of overheads of the plurality of task types. One or more of the estimates of overheads of the plurality of task types are varied.

摘要翻译： 在包括多个计算设备的计算机系统中，其中所述多个计算设备处理多个任务，并且每个任务具有任务类型，用于确定与任务类型相关联的开销的方法包括以下步骤。 估计多个任务类型的开销。 选择多个计算设备中的一个来执行多个任务中的一个，其中所述选择包括从分配给所述多个计算设备的至少一部分的任务估计所述多个计算设备的至少一部分上的负载，以及 多个任务类型的开销的估计。 改变多个任务类型的开销的估计中的一个或多个。

公开(公告)号：US08316109B2

公开(公告)日：2012-11-20

申请号：US13045392

申请日：2011-03-10

申请人： Douglas M. Freimuth ,  Elbert C. Hu ,  Ronald Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John M. Tracey

发明人： Douglas M. Freimuth ,  Elbert C. Hu ,  Ronald Mraz ,  Erich M. Nahum ,  Prashant Pradhan ,  Sambit Sahu ,  John M. Tracey

IPC分类号： G06F15/16

CPC分类号： G06F13/30

摘要： A number of improvements in network adapters that offload protocol processing from the host processor are provided. Specifically, mechanisms for handling memory management and optimization within a system utilizing an offload network adapter are provided. The memory management mechanism permits both buffered sending and receiving of data as well as zero-copy sending and receiving of data. In addition, the memory management mechanism permits grouping of DMA buffers that can be shared among specified connections based on any number of attributes. The memory management mechanism further permits partial send and receive buffer operation, delaying of DMA requests so that they may be communicated to the host system in bulk, and expedited transfer of data to the host system.

摘要翻译： 提供了从主机处理器卸载协议处理的网络适配器的许多改进。 具体地说，提供了利用卸载网络适配器在系统内处理存储器管理和优化的机制。 存储器管理机制允许数据的缓冲发送和接收以及零拷贝发送和接收数据。 此外，存储器管理机制允许基于任何数量的属性对可以在指定连接之间共享的DMA缓冲器进行分组。 存储器管理机制进一步允许部分发送和接收缓冲器操作，延迟DMA请求，使得它们可以批量地传送到主机系统，并且将数据快速传送到主机系统。