公开(公告)号：US07289509B2

公开(公告)日：2007-10-30

申请号：US10076340

申请日：2002-02-14

申请人： Deanna Lynn Quigg Brown ,  Vinit Jain ,  Vasu Vallabhaneni ,  Venkat Venkatsubra

发明人： Deanna Lynn Quigg Brown ,  Vinit Jain ,  Vasu Vallabhaneni ,  Venkat Venkatsubra

IPC分类号： H04L12/28 ,  H04L12/56

CPC分类号： H04L45/00 ,  H04L29/06 ,  H04L45/24 ,  H04L47/10 ,  H04L47/34 ,  H04L69/14 ,  H04L69/16 ,  H04L69/161 ,  H04L69/166

摘要： A method, system and apparatus for improving data transfer rate over a network are provided. When a piece of data is to be transmitted to a target system, it is divided into a number of packets and a determination is made as to whether the number of packets exceeds a threshold number. If so, the packets are transferred in parallel over a plurality of network connections. The ideal network connections are TCP/IP connections. Thus, each packet has an IP header, which contains an IP identification field. An indicium is placed into the IP identification field for proper reconstruction of the data by the target system.

摘要翻译： 提供了一种用于提高网络上的数据传输速率的方法，系统和装置。 当一条数据要发送到目标系统时，将其分成多个分组，并确定分组数是否超过阈值数。 如果是这样，则分组在多个网络连接上并行传送。 理想的网络连接是TCP / IP连接。 因此，每个分组具有包含IP标识字段的IP报头。 将标记放入IP识别字段，以便目标系统正确重建数据。

公开(公告)号：US20060133379A1

公开(公告)日：2006-06-22

申请号：US11014540

申请日：2004-12-16

申请人： Uma Krishnan ,  Venkat Venkatsubra ,  Vinit Jain

发明人： Uma Krishnan ,  Venkat Venkatsubra ,  Vinit Jain

IPC分类号： H04L1/00

CPC分类号： H04L1/16 ,  H04L1/1809 ,  H04L69/16 ,  H04L69/163 ,  H04W8/26 ,  H04W28/06 ,  H04W80/06

摘要： A TCP/IP retransmission frame contains multiple TCP segments containing non-contiguous missing packets from a data communication. A TCP option in the first segment in the frame indicates the presence of multiple TCP payloads in the frame. Each TCP payload has a corresponding prepended TCP header. An offset address specified in each TCP header within the frame identifies the location of the next TCP header in the frame. Network throughput performance is enhanced by permitting multiple non-contiguous packets to be retransmitted in a single IP packet.

公开(公告)号：US07050393B2

公开(公告)日：2006-05-23

申请号：US09852826

申请日：2001-05-10

申请人： Deanna Lynn Quigg Brown ,  Vinit Jain ,  Venkat Venkatsubra

发明人： Deanna Lynn Quigg Brown ,  Vinit Jain ,  Venkat Venkatsubra

IPC分类号： H04L12/28 ,  H04J1/16 ,  G08C15/00 ,  G06F11/00

CPC分类号： H04L47/10 ,  H04L47/26

摘要： A method, system, and product are described for alleviating router congestion when the router is processing packets transmitted by computer systems having a congestion notification capability. The router marks packets transmitted between a first computer system and a second computer system when the first and second computer systems have the congestion notification capability. The marking indicates that the router is moderately congested. A marked packet is then received by the first computer system. The router then drops packets transmitted by the first computer system when the packets were transmitted subsequently to the receipt of the marked packet.

摘要翻译： 描述了一种方法，系统和产品，用于在路由器处理具有拥塞通知能力的计算机系统发送的分组时减轻路由器拥塞。 当第一和第二计算机系统具有拥塞通知能力时，路由器标记在第一计算机系统和第二计算机系统之间传输的分组。 标记表示路由器中等拥塞。 然后，标记的分组被第一计算机系统接收。 然后路由器丢弃第一计算机系统发送的数据包，随后在接收到标记的数据包之后发送数据包。

公开(公告)号：US20060107324A1

公开(公告)日：2006-05-18

申请号：US10992514

申请日：2004-11-18

申请人： Radhika Chirra ,  Ranadip Das ,  Vinit Jain ,  Venkat Venkatsubra

发明人： Radhika Chirra ,  Ranadip Das ,  Vinit Jain ,  Venkat Venkatsubra

IPC分类号： G06F12/14

CPC分类号： H04L63/1458 ,  H04L69/16 ,  H04L69/163 ,  H04L2463/141

摘要： An improved method, apparatus, and computer instructions for preventing denial of service attacks on persistent connections. A synchronize packet is received. In response to receiving the synchronize packet, a state of the persistent connection is identified. An action on the synchronize packet is deferred until a subsequent communication with a peer to the persistent connection.

摘要翻译： 用于防止对持久连接的拒绝服务攻击的改进的方法，装置和计算机指令。 接收到同步数据包。 响应于接收到同步分组，识别持续连接的状态。 对同步数据包的操作被延迟到与对等体的后续通信到持久连接。

公开(公告)号：US20060031409A1

公开(公告)日：2006-02-09

申请号：US10912498

申请日：2004-08-05

申请人： Vinit Jain ,  Kavitha Baratakke ,  Jeffrey Messing ,  Venkat Venkatsubra

发明人： Vinit Jain ,  Kavitha Baratakke ,  Jeffrey Messing ,  Venkat Venkatsubra

IPC分类号： G06F15/16

CPC分类号： H04L67/2842 ,  H04L49/90 ,  H04L49/9036 ,  H04L69/16 ,  H04L69/161 ,  H04L69/32

摘要： A method, system, and computer program product in a computer-readable medium for delivering data, received from a network, to a storage buffer assigned to an application is proposed. An application designates a communication buffer within a local data processing system for buffering data communicated with an application. The local data processing system reports to a network interface of the local data processing system a memory address of the designated communication buffer, and the data processing system creates a cookie containing the memory address. The data processing system then sends the cookie form the local data processing system to a remote data processing system, such that the remote data processing system may address data directly to the designated communication buffer.

公开(公告)号：US20050091307A1

公开(公告)日：2005-04-28

申请号：US10692124

申请日：2003-10-23

申请人： Venkat Venkatsubra ,  Vinit Jain

发明人： Venkat Venkatsubra ,  Vinit Jain

IPC分类号： H04L29/06 ,  G06F15/16

CPC分类号： H04L65/4092 ,  H04L29/06027 ,  H04L65/608

摘要： A graphic user interface of a network client (106) includes a stock ticker from a stock server (110) and a news sidebar from a news server (112) over the IP addresses advertised to the client (106) in its list of mutli-homed addresses from the video server (104) specified under Stream Control Transmission Protocol (SCTP). The client accepts real-time data from the stock exchange server and the news agency server on the multi-homed IP addresses designated in the association with the video server (104) without knowing that the data is coming from a different source than the video server (104). The real-time data feeds from the video, stock and news servers are aggregated on the client (106) with enhanced speed because the feeds come directly to the client and not via the video server. The operating systems of the home server and remote servers utilize SCTP and specialized commands to implement the enhanced speed of real-time data aggregation being streamed to network clients without requiring modifications to existing client systems.

摘要翻译： 网络客户端（106）的图形用户界面包括来自股票服务器（110）的股票行情和来自新闻服务器（112）的新闻侧栏，通过在其列表中通告给客户端（106）的IP地址， 来自在流控制传输协议（SCTP）下指定的视频服务器（104）的归属地址。 客户端接收来自证券交易所服务器和通讯社服务器的与视频服务器（104）相关联指定的多归属IP地址的实时数据，而不知道数据来自与视频服务器不同的源 （104）。 来自视频，股票和新闻服务器的实时数据馈送以高速度聚合在客户端（106）上，因为馈送直接进入客户端而不是经由视频服务器。 家庭服务器和远程服务器的操作系统使用SCTP和专用命令来实现流式传输到网络客户端的实时数据聚合的增强速度，而不需要修改现有的客户端系统。

公开(公告)号：US20050086359A1

公开(公告)日：2005-04-21

申请号：US10687236

申请日：2003-10-16

申请人： Dwip Banerjee ,  Kavitha Baratakke ,  Vinit Jain ,  Venkat Venkatsubra

发明人： Dwip Banerjee ,  Kavitha Baratakke ,  Vinit Jain ,  Venkat Venkatsubra

IPC分类号： G06F9/46 ,  G06F9/50 ,  G06F15/16

CPC分类号： G06F9/5027 ,  G06F9/544 ,  G06F2209/5018 ,  G06F2209/508

摘要： A method, system, and program for monitoring thread usage to dynamically control a thread pool are provided. An application running on the server system invokes a listener thread on a listener socket for receiving client requests at the server system and passing the client requests to one of multiple threads waiting in a thread pool. Additionally, the application sends an ioctl call in blocking mode on the listener thread. A TCP layer within the server system detects the listener thread in blocking mode and monitors a thread count of at least one of a number of incoming requests waiting to be processed and a number of said plurality of threads remaining idle in the thread pool over a sample period. Once the TCP layer detects a thread usage event, the ioctl call is returned indicating the thread usage event with the thread count, such that a number of threads in the thread pool may be dynamically adjusted to handle the thread count.

摘要翻译： 提供了一种用于监视线程使用以动态控制线程池的方法，系统和程序。 在服务器系统上运行的应用程序调用侦听器套接字上的侦听器线程，以便在服务器系统上接收客户端请求，并将客户端请求传递给等待线程池的多个线程之一。 此外，应用程序在侦听器线程上以阻塞模式发送ioctl调用。 服务器系统内的TCP层以阻塞模式检测侦听器线程，并且监视等待被处理的多个传入请求中的至少一个的线程计数，并且线程池中的多个线程在一个样本上保持空闲 期。 一旦TCP层检测到线程使用事件，将返回指示线程计数的线程使用事件的ioctl调用，以使线程池中的线程数可以被动态调整以处理线程计数。

公开(公告)号：US20050036511A1

公开(公告)日：2005-02-17

申请号：US10640818

申请日：2003-08-14

申请人： Kavitha Baratakke ,  Vinit Jain ,  Vasu Vallabhaneni ,  Venkat Venkatsubra

发明人： Kavitha Baratakke ,  Vinit Jain ,  Vasu Vallabhaneni ,  Venkat Venkatsubra

IPC分类号： H04L29/06 ,  H04J3/16

CPC分类号： H04L29/06 ,  H04L69/16 ,  H04L69/161 ,  H04L69/163

摘要： TCP congestion avoidance is implemented upon retransmission of a packet and is reverted back to the original congestion state upon receipt of an early acknowledgement (ACK), indicating reordering of packets, thereby eliminating a needless restriction on TCP bandwidth. Upon receiving an ACK to a retransmitted packet, it is determined if the ACK resulted from receipt of the original reordered packet or the retransmitted packet, based on the arrival time of the ACK at the sender. If the round-trip-time (RTT) for the retransmitted packet is much lower than the average or current calculated RTT for the network link between sender and receiver, then the retransmission occurred as a result of a reordering event, and the congestion window is restored back to its value prior to the retransmission, thereby permitting the network link to continue operating at its original increased throughput.

摘要翻译： TCP拥塞避免在分组重传时实现，并且在接收到指示重新排序分组的早期确认（ACK）时被恢复到原始拥塞状态，从而消除对TCP带宽的不必要的限制。 在向重传的分组接收到ACK时，基于ACK在发送方的到达时间确定ACK是否从接收到原始重新排序的分组或重传分组而产生。 如果重发分组的往返时间（RTT）远低于发送方和接收方之间的网络链路的平均或当前计算的RTT，则重新发送作为重排序事件的结果，拥塞窗口为 在重传之前恢复其值，从而允许网络链路以其原始增加的吞吐量继续运行。

公开(公告)号：US20050036452A1

公开(公告)日：2005-02-17

申请号：US10640819

申请日：2003-08-14

申请人： Dwip Banerjee ,  Vinit Jain ,  Vasu Vallabhaneni ,  Venkat Venkatsubra

发明人： Dwip Banerjee ,  Vinit Jain ,  Vasu Vallabhaneni ,  Venkat Venkatsubra

IPC分类号： H04J1/16 ,  H04L1/00 ,  H04L1/16 ,  H04L1/18 ,  H04L1/24 ,  H04L12/56 ,  H04L29/06 ,  H04L29/14

CPC分类号： H04L1/0061 ,  H04L1/0072 ,  H04L1/08 ,  H04L1/16 ,  H04L1/1809 ,  H04L45/00 ,  H04L45/22 ,  H04L45/28 ,  H04L69/16 ,  H04L69/161 ,  H04L69/163 ,  H04L69/22 ,  H04L69/40

摘要： A receiving host in a TCP/IP network sends an acknowledgment indicating a received data packet is corrupt. The sending host will begin transmitting with a new field set in the IP header called a check-TCP-checksum bit, thereby requesting that all routers in the TCP/IP network perform a checksum on the entire received packet. Routers in the TCP/IP network will perform a complete checksum on an entire packet with the check-TCP-checksum bit set, and not just on the IP header. The routers continuously monitor the ratio of corrupt packets received on a particular port that fail the entire packet checksum to the total number of packets received on that port. If the ratio of corrupt-to-received packets exceeds a corruption threshold, the router assumes that the associated link is causing data corruption and issues a routing update indicating that the link is bad and should be avoided. Once the retransmission rate between the sender and receiver drops below a threshold level, the bad link has been detected and avoided within the TCP/IP network and the check-TCP-checksum option in the IP header is no longer set in data packets transmitted to the receiver host.

摘要翻译： TCP / IP网络中的接收主机发送指示接收到的数据包已损坏的确认。 发送主机将开始使用称为检查TCP校验和位的IP报头中设置的新字段进行发送，从而请求TCP / IP网络中的所有路由器对整个接收到的数据包执行校验和。 TCP / IP网络中的路由器将对整个数据包执行完整的校验和，同时检查TCP-checksum位，而不仅仅是在IP报头上。 路由器连续监视在特定端口上接收的损坏数据包的总数与在该端口上接收的数据包的总数之间的比例。 如果损坏的数据包的比例超过了破坏阈值，则路由器假设相关联的链路导致数据损坏，并发出路由更新，指示链路不良，应该避免。 一旦发送方和接收方之间的重传速率下降到阈值以下，就会在TCP / IP网络内检测并避免不良链路，并且IP报头中的check-TCP-checksum选项不再被设置在发送到 接收主机。

公开(公告)号：US09092274B2

公开(公告)日：2015-07-28

申请号：US13313534

申请日：2011-12-07

申请人： Omar Cardona ,  Vinit Jain ,  Renato J. Recio ,  Rakesh Sharma

发明人： Omar Cardona ,  Vinit Jain ,  Renato J. Recio ,  Rakesh Sharma

IPC分类号： G06F15/16 ,  G06F9/54 ,  G06F9/455

CPC分类号： G06F9/54 ,  G06F9/45558 ,  G06F2009/45595

摘要： An approach is provided in which a hardware accelerated bridge executing on a network adapter receives an ingress data packet. The data packet includes a destination MAC address that corresponds to a virtual machine, which interfaces to a software bridge executing on a hypervisor. The hardware accelerated bridge identifies a software bridge table entry that includes the destination MAC address and a virtual function identifier, which identifies a virtual function corresponding to the software bridge. In turn, the hardware accelerated bridge sends the data packet from the hardware accelerated bridge to the software bridge through the identified virtual function.

摘要翻译： 提供了一种方法，其中在网络适配器上执行的硬件加速桥接收入入数据分组。 数据包包括与虚拟机相对应的目的地MAC地址，该目的MAC地址与在管理程序上执行的软件桥接口相连接。 硬件加速桥标识一个包含目标MAC地址和虚拟功能标识符的软件桥表项，它标识与软件桥相对应的虚拟功能。 反过来，硬件加速网桥通过识别的虚拟功能将数据包从硬件加速桥发送到软件桥。