公开(公告)号：US20050270972A1

公开(公告)日：2005-12-08

申请号：US11141705

申请日：2005-05-31

Applicant: Muralidharan Kodialam ,  Tirunell Lakshman ,  Sudipta Sengupta

Inventor： Muralidharan Kodialam ,  Tirunell Lakshman ,  Sudipta Sengupta

IPC: H04L12/56

CPC classification number: H04L45/12 ,  H04L45/04 ,  H04L45/24 ,  H04L47/125

Abstract: A method for supporting recovery from failure of a path in a network of nodes interconnected by links comprises: (a) selecting an intermediate node between an ingress point and an egress point of the network, wherein the intermediate node minimizes the sum of (i) a capacity constraint between the ingress point and the intermediate node and (ii) a capacity constraint between the intermediate node and the egress point; wherein the selection identifies a first link-disjoint path set between the ingress point and the intermediate node, and a second link-disjoint path set between the intermediate node and the egress point, each link-disjoint path set comprising a backup path and at least one primary path; (b) implementing, during a first routing phase, a first routing method for routing a fraction of a service level between the ingress point and the intermediate node along each of the one or more primary paths of the first link-disjoint path set; and (c) implementing, during a second routing phase, a second routing method for routing a fraction of the service level between the intermediate node and the egress point along each of the one or more primary paths of the second link-disjoint path set.

Abstract translation: 一种用于支持通过链路互连的节点网络中的路径故障恢复的方法包括：（a）在所述网络的入口点和出口点之间选择中间节点，其中所述中间节点将（i） 入口点和中间节点之间的容量约束和（ii）中间节点与出口点之间的容量约束; 其中所述选择识别在所述入口点和所述中间节点之间设置的第一链接 - 不相交路径，以及在所述中间节点和所述出口点之间设置的第二链路 - 不相交路径，每个链路 - 不相交路径集合包括备用路径并且至少 一条主路 （b）在第一路由阶段期间实施第一路由方法，用于沿着所述第一链路 - 不相交路径集合的所述一个或多个主路径中的每一个路由在所述入口点和所述中间节点之间的服务级别的一部分; 以及（c）在第二路由阶段期间实施第二路由选择方法，用于沿着所述第二链路 - 不相交路径集合的所述一个或多个主路径中的每一个路由在所述中间节点和所述出口点之间的所述服务级别的一部分。

公开(公告)号：US08996683B2

公开(公告)日：2015-03-31

申请号：US12410697

申请日：2009-03-25

Applicant: David A. Maltz ,  Albert G. Greenberg ,  Parveen K. Patel ,  Sudipta Sengupta ,  Parantap Lahiri

Inventor： David A. Maltz ,  Albert G. Greenberg ,  Parveen K. Patel ,  Sudipta Sengupta ,  Parantap Lahiri

IPC: G06F15/173 ,  G06F15/16 ,  H04L29/12 ,  H04L12/707 ,  H04L12/803

CPC classification number: H04L29/12056 ,  H04L29/12028 ,  H04L45/24 ,  H04L47/125 ,  H04L61/103 ,  H04L61/1505

Abstract: A method of networking a plurality of servers together within a data center is disclosed. The method includes the step of addressing a data packet for delivery to a destination server by providing the destination server address as a flat address. The method further includes the steps of obtaining routing information required to route the packet to the destination server. This routing information may be obtained from a directory service servicing the plurality of servers. Once the routing information is obtained, the data packet may be routed to the destination server according to the flat address of the destination server and routing information obtained from the directory service.

Abstract translation: 公开了一种在数据中心内组合多个服务器的方法。 该方法包括通过提供目的地服务器地址作为平坦地址来寻址用于传送到目的地服务器的数据分组的步骤。 该方法还包括步骤：获得将分组路由到目的地服务器所需的路由信息​​。 可以从服务于多个服务器的目录服务获得该路由信息。 一旦获得了路由信息，数据包可以根据目的地服务器的平面地址和从目录服务获得的路由信息​​被路由到目的地服务器。

公开(公告)号：US20130346672A1

公开(公告)日：2013-12-26

申请号：US13531455

申请日：2012-06-22

Applicant: Sudipta Sengupta ,  Jin Li ,  Cheng Huang ,  Timothy Andrew Pritchett ,  Christopher Broder Wilson

Inventor： Sudipta Sengupta ,  Jin Li ,  Cheng Huang ,  Timothy Andrew Pritchett ,  Christopher Broder Wilson

IPC: G06F12/02 ,  G06F12/08

CPC classification number: G06F12/0871 ,  G06F12/0897 ,  G06F12/12

Abstract: The subject disclosure is directed towards a multi-tiered cache having cache tiers with different access properties. Objects are written to a selected a tier of the cache based upon object-related properties and/or cache-related properties. In one aspect, objects are stored in an active log among a plurality of logs. The active log is sealed upon reaching a target size, with a new active log opened. Garbage collecting is performed on a sealed log, such as the sealed log with the most garbage therein.

Abstract translation: 主题公开针对具有不同访问属性的高速缓存层的多层缓存。 基于对象相关属性和/或缓存相关属性，将对象写入高速缓存的选定层。 在一个方面，对象被存储在多个日志中的活动日志中。 活动日志在达到目标大小时被打开，新的活动日志打开。 垃圾收集是在密封的日志上进行的，例如其中垃圾最多的密封日志。

公开(公告)号：US20120102298A1

公开(公告)日：2012-04-26

申请号：US12908153

申请日：2010-10-20

Applicant: Sudipta Sengupta ,  Biplob Kumar Debnath ,  Jin Li

Inventor： Sudipta Sengupta ,  Biplob Kumar Debnath ,  Jin Li

IPC: G06F12/10 ,  G06F12/00

Abstract: Described is using flash memory (or other secondary storage), RAM-based data structures and mechanisms to access key-value pairs stored in the flash memory using only a low RAM space footprint. A mapping (e.g. hash) function maps key-value pairs to a slot in a RAM-based index. The slot includes a pointer that points to a bucket of records on flash memory that each had keys that mapped to the slot. The bucket of records is arranged as a linear-chained linked list, e.g., with pointers from the most-recently written record to the earliest written record. Also described are compacting non-contiguous records of a bucket onto a single flash page, and garbage collection. Still further described is load balancing to reduce variation in bucket sizes, using a bloom filter per slot to avoid unnecessary searching, and splitting a slot into sub-slots.

Abstract translation: 描述的是使用闪速存储器（或其他辅助存储器），基于RAM的数据结构和机制来访问存储在闪存中的键值对，仅使用低的RAM空间。 映射（例如散列）功能将键值对映射到基于RAM的索引中的时隙。 插槽包括一个指向闪存中的一桶记录的指针，每个记录都具有映射到插槽的键。 记录的桶被布置为线性链接的链表，例如具有从最近写入的记录到最早的书面记录的指针。 还描述了将桶的不连续记录压缩到单个闪存页面上以及垃圾收集。 还进一步描述的是负载平衡以减少桶大小的变化，使用每时隙的布隆过滤器来避免不必要的搜索，并且将时隙分割成子时隙。

公开(公告)号：US08160063B2

公开(公告)日：2012-04-17

申请号：US12410745

申请日：2009-03-25

Applicant: David A. Maltz ,  Albert G. Greenberg ,  Parveen K. Patel ,  Sudipta Sengupta ,  Parantap Lahiri

Inventor： David A. Maltz ,  Albert G. Greenberg ,  Parveen K. Patel ,  Sudipta Sengupta ,  Parantap Lahiri

IPC: H04L12/56 ,  H04J1/16

CPC classification number: H04L45/24 ,  H04L45/02 ,  H04L45/28

Abstract: A system for commoditizing data center networking is disclosed. The system includes an interconnection topology for a data center having a plurality of servers and a plurality of nodes of a network in the data center through which data packets may be routed. The system uses a routing scheme where the routing is oblivious to the traffic pattern between nodes in the network, and wherein the interconnection topology contains a plurality of paths between one or more servers. The multipath routing may be Valiant load balancing. It disaggregates the function of load balancing into a group of regular servers, with the result that load balancing server hardware can be distributed amongst racks in the data center leading to greater agility and less fragmentation. The architecture creates a huge, flexible switching domain, supporting any server/any service, full mesh agility, and unregimented server capacity at low cost.

Abstract translation: 公开了一种用于商业化数据中心网络的系统。 该系统包括用于具有数据中心的多个服务器和多个节点的数据中心的互连拓扑，通过该数据中心可以路由数据分组。 系统使用路由方案，其中路由忽略网络中的节点之间的流量模式，并且其中互连拓扑在一个或多个服务器之间包含多个路径。 多路径路由可能是Valiant负载平衡。 它将负载平衡的功能分解为一组常规服务器，从而可以在数据中心的机架之间分配负载平衡服务器硬件，从而实现更高的灵活性和更少的分段。 该架构创建了一个巨大的，灵活的交换域，以低成本支持任何服务器/任何服务，全面的网格敏捷性和未注册的服务器容量。

公开(公告)号：US08082358B2

公开(公告)日：2011-12-20

申请号：US12242913

申请日：2008-09-30

Applicant: Jiajun Wang ,  Cheng Huang ,  Sudipta Sengupta ,  Jin Li

Inventor： Jiajun Wang ,  Cheng Huang ,  Sudipta Sengupta ,  Jin Li

IPC: G06F15/16

CPC classification number: H04L47/19 ,  H04L47/10 ,  H04L47/12 ,  H04L47/25 ,  H04L67/104

Abstract: An ISP-friendly rate allocation system and method that reduces network traffic across ISP boundaries in a peer-to-peer (P2P) network, Embodiments of the system and method continuously solve a global optimization problem and dictate accordingly how much bandwidth is allocated on each connection. Embodiments of the system and method minimize load on a server in communication with the P2P network, minimize ISP-unfriendly traffic while keeping the minimum server load unaffected, and maximize peer prefetching. Two different techniques are used to compute rate allocation, including a utility function optimization technique and a minimum cost flow formulation technique. The utility function optimization technique constructs a utility function and optimizes that utility function. The minimum cost flow formulation technique generates a minimum cost flow formulation using a bipartite graph have a vertices set and an edges set. A distributed minimum cost flow formulation is solved using Lagrangian multipliers.

Abstract translation: 一种ISP友好的速率分配系统和方法，可减少对等（P2P）网络中ISP边界的网络流量，系统和方法的实施方案不断解决全局优化问题，并据此规定每个网络上分配多少带宽 连接。 系统和方法的实施例最小化与P2P网络通信的服务器上的负载，从而最小化ISP不友好的业务，同时保持最小服务器负载不受影响，并最大化对等体预取。 使用两种不同的技术来计算速率分配，包括效用函数优化技术和最低成本流量制定技术。 效用函数优化技术构建效用函数并优化该效用函数。 最小成本流量制定技术使用具有顶点集合和边集合的二分图生成最小成本流程公式。 使用拉格朗日乘数解决分布式最小成本流公式。

公开(公告)号：US20110276780A1

公开(公告)日：2011-11-10

申请号：US12979644

申请日：2010-12-28

Applicant: Sudipta Sengupta ,  Biplob Debnath ,  Jin Li ,  Ronakkumar N. Desai ,  Paul Adrian Oltean

Inventor： Sudipta Sengupta ,  Biplob Debnath ,  Jin Li ,  Ronakkumar N. Desai ,  Paul Adrian Oltean

IPC: G06F12/08

CPC classification number: G06F12/0862 ,  G06F12/0866 ,  G06F17/30159

Abstract: The subject disclosure is directed towards a data deduplication technology in which a hash index service's index maintains a hash index in a secondary storage device such as a hard drive, along with a compact index table and look-ahead cache in RAM that operate to reduce the I/O to access the secondary storage device during deduplication operations. Also described is a session cache for maintaining data during a deduplication session, and encoding of a read-only compact index table for efficiency.

公开(公告)号：US20110276744A1

公开(公告)日：2011-11-10

申请号：US12773859

申请日：2010-05-05

Applicant: Sudipta Sengupta ,  Biplob Kumar Debnath ,  Jin Li

Inventor： Sudipta Sengupta ,  Biplob Kumar Debnath ,  Jin Li

IPC: G06F12/02 ,  G06F12/10 ,  G06F12/00

CPC classification number: G06F12/0246 ,  G06F12/0866 ,  G06F12/0897 ,  G06F12/123 ,  G06F2212/222 ,  G11C7/1072

Abstract: Described is using flash memory, RAM-based data structures and mechanisms to provide a flash store for caching data items (e.g., key-value pairs) in flash pages. A RAM-based index maps data items to flash pages, and a RAM-based write buffer maintains data items to be written to the flash store, e.g., when a full page can be written. A recycle mechanism makes used pages in the flash store available by destaging a data item to a hard disk or reinserting it into the write buffer, based on its access pattern. The flash store may be used in a data deduplication system, in which the data items comprise chunk-identifier, metadata pairs, in which each chunk-identifier corresponds to a hash of a chunk of data that indicates. The RAM and flash are accessed with the chunk-identifier (e.g., as a key) to determine whether a chunk is a new chunk or a duplicate.

Abstract translation: 描述的是使用闪存，基于RAM的数据结构和机制来提供用于在闪存页中缓存数据项（例如键值对）的闪存。 基于RAM的索引将数据项映射到闪存页面，并且基于RAM的写入缓冲器保持要写入闪存存储器的数据项目，例如当可以写入全页时。 回收机制使得通过将数据项降级到硬盘或将其重新插入到写入缓冲器中，基于其访问模式，可用于闪存存储器中的使用页面。 闪存存储器可以用在数据重复数据删除系统中，其中数据项包括块标识符，元数据对，其中每个块标识符对应于指示的数据块的散列。 使用块标识符（例如，作为密钥）来访问RAM和闪存，以确定块是新的块还是重复的。

公开(公告)号：US07957266B2

公开(公告)日：2011-06-07

申请号：US11106410

申请日：2005-04-14

Applicant: Muralidharan S. Kodialam ,  Tirunell V. Lakshman ,  Sudipta Sengupta

Inventor： Muralidharan S. Kodialam ,  Tirunell V. Lakshman ,  Sudipta Sengupta

IPC: H04L12/26

CPC classification number: H04L45/12 ,  H04L45/04 ,  H04L45/24 ,  H04L47/125

Abstract: A scheme for routing packets of traffic to their destination after ensuring that they pass through one or more pre-determined intermediate nodes, thereby permitting all permissible traffic patterns to be handled without knowledge of the traffic matrix, subject to edge-link capacity constraints. In one embodiment, a request for a path with a service demand for routing data between the ingress point and the egress point is received. A set of two or more intermediate nodes between the ingress point and the egress point is selected. Based on a bandwidth of the network, respective fractions of the data to send from the ingress point to each node of the set of intermediate nodes are determined. The data is routed in the determined respective fractions from the ingress point to each node of the set of intermediate nodes, and routed from each node of the set of intermediate nodes to the egress point.

Abstract translation: 一种用于在确保它们通过一个或多个预定中间节点之后将流量分组路由到其目的地的方案，从而允许所有允许的业务模式在不了解业务矩阵的情况下被处理，而不受边缘链路容量限制。 在一个实施例中，接收到对于在入口点和出口点之间路由数据的服务需求的路径的请求。 选择入口点和出口点之间的两个或多个中间节点的集合。 基于网络的带宽，确定从入口点向中间节点集合中的每个节点发送的数据的各个分数。 将数据以确定的各个分数从入口点路由到中间节点集合中的每个节点，并从中间节点集合的每个节点路由到出口点。

公开(公告)号：US07835271B2

公开(公告)日：2010-11-16

申请号：US11321549

申请日：2005-12-29

Applicant: Mansoor Ali Khan Alicherry ,  Peter Hessler ,  Viswanath Poosala ,  Walter Rothkegel ,  Sudipta Sengupta

Inventor： Mansoor Ali Khan Alicherry ,  Peter Hessler ,  Viswanath Poosala ,  Walter Rothkegel ,  Sudipta Sengupta

IPC: G01R31/08

CPC classification number: H04J14/0227 ,  H04J3/085 ,  H04J14/0284 ,  H04J14/0291 ,  H04J2203/006 ,  H04L12/437

Abstract: Improved p-cycle restoration techniques using a signaling protocol are disclosed. For example, a technique for use in at least one node of a data communication network for recovering from a failure, wherein the data communication network includes multiple nodes and multiple links for connecting the multiple nodes, comprises the following steps/operations. Notification of the failure is obtained at the at least one node. A determination is made whether the failure is a single link failure or one of a node failure and a multiple link failure. A pre-configured protection cycle (p-cycle) plan is implemented when the failure is a single link failure but not when the failure is one of a node failure and a multiple link failure, such that two independent paths in the network are not connected when implementing the pre-configured protection cycle plan. Implementation of the pre-configured protection cycle plan may further comprise the node sending at least one message to another node in the data communication network and/or receiving at least one message from another node in the data communication network.

Abstract translation: 公开了使用信令协议的改进的p循环恢复技术。 例如，一种在数据通信网络的至少一个节点中用于从故障中恢复的技术，其中数据通信网络包括多个节点和用于连接多个节点的多个链路，包括以下步骤/操作。 在至少一个节点处获得故障的通知。 确定故障是单个链路故障还是节点故障和多个链路故障之一。 当故障是单个链路故障时，实施预配置的保护周期（p周期）计划，但是当故障是节点故障和多个链路故障之一时，则不实现，这样网络中的两个独立路径不连接 在实施预先配置的保护周期计划时。 预先配置的保护周期计划的实施还可以包括节点向数据通信网络中的另一节点发送至少一个消息和/或从数据通信网络中的另一个节点接收至少一个消息。