公开(公告)号：US08037349B2

公开(公告)日：2011-10-11

申请号：US12549847

申请日：2009-08-28

申请人： Nagapramod S. Mandagere ,  Seshashayee S. Murthy ,  Mark A. Smith ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Nagapramod S. Mandagere ,  Seshashayee S. Murthy ,  Mark A. Smith ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F11/00

CPC分类号： G06F11/1435 ,  G06F11/1453

摘要： A system and associated method for replicating data based on capacity optimization. A local node receives the data associated with a key. The local node within a local domain communicates with nodes of remote domains in a system through a communication network. Each domain has its own distributed hash table that partitions key space and assigns a certain key range to an owner node within the domain. For new data, the local node queries owner nodes of domains in the system progressively from the local domain to remote domains for a duplicate of the new data. Depending on a result returned by owner nodes and factors for replication strategies, the local node determines a replication strategy and records the new data in the local node pursuant to the replication strategy.

摘要翻译： 一种基于容量优化来复制数据的系统和相关方法。 本地节点接收与密钥相关联的数据。 本地域内的本地节点通过通信网络与系统中的远程域的节点进行通信。 每个域都有自己的分布式哈希表，用于分区密钥空间，并为域内的所有者节点分配一定的密钥范围。 对于新数据，本地节点从本地域向远程域逐行查询系统中的域的所有者节点，以获取新数据的副本。 根据所有者节点返回的结果和复制策略的因素，本地节点确定复制策略，并根据复制策略将新数据记录在本地节点中。

公开(公告)号：US08775376B2

公开(公告)日：2014-07-08

申请号：US13172930

申请日：2011-06-30

申请人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F7/00 ,  G06F17/00

CPC分类号： G06F11/1458 ,  G06F11/1448 ,  G06F11/1464

摘要： Embodiments of the present invention provide a hybrid (e.g., local and remote) approach for data backup in a networked computing environment (e.g., a cloud computing environment). In a typical embodiment, a set of storage configuration parameters corresponding to a set of data to be backed up is received and stored in a computer data structure. The set of storage configuration parameters can comprise at least one of the following: a recovery time objective (RTO), a recovery point objective (RPO), and a desired type of protection for the set of data. Regardless, the set of data is compared to previously stored data to identify at least one of the following: portions of the set of data that have commonality with the previously stored data; and portions of the set of data that are unique to the set of data (i.e., not in common with any of the previously stored data). The above-described process is referred to herein as “de-duplication”. A storage solution is then determined based on the set of storage configuration parameters. In general, the storage solution identifies at least one local storage resource and at least one remote storage resource (e.g., a cloud storage resource) for backing up the portions of the set of data that are unique to the set of data. Once the storage solution has been determined, the unique portions of the set of data will be stored in accordance therewith.

摘要翻译： 本发明的实施例提供了一种用于网络计算环境（例如，云计算环境）中的数据备份的混合（例如，本地和远程）方法。 在典型的实施例中，接收一组对应于要备份的数据的存储配置参数并将其存储在计算机数据结构中。 该组存储配置参数可以包括以下中的至少一个：恢复时间目标（RTO），恢复点目标（RPO）以及该组数据的期望类型的保护。 无论如何，将该组数据与先前存储的数据进行比较，以识别以下中的至少一个：与先前存储的数据具有共同性的数据集合中的部分; 以及数据集合中对数据集合唯一的部分（即，与先前存储的任何数据不共同的）。 上述处理在本文中被称为“重复数据删除”。 然后基于存储配置参数的集合来确定存储解决方案。 通常，存储解决方案识别至少一个本地存储资源和至少一个远程存储资源（例如，云存储资源），用于备份该组数据的唯一的数据组的部分。 一旦确定了存储解决方案，则将根据其存储该组数据的唯一部分。

公开(公告)号：US08781800B2

公开(公告)日：2014-07-15

申请号：US13286490

申请日：2011-11-01

申请人： Kavita Chavda ,  Eric W. Davis Rozier ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Kavita Chavda ,  Eric W. Davis Rozier ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F17/10

CPC分类号： G06F7/00 ,  G06F17/30159

摘要： Embodiments of the present invention provide an approach that utilizes discrete event simulation to quantitatively analyze the reliability of a modeled de-duplication system in a computer storage environment. In addition, the approach described herein can perform such an analysis on systems having heterogeneous data stored on heterogeneous storage systems in the presence of primary faults and their secondary effects due to de-duplication. In a typical embodiment, data de-duplication parameters and a hardware configuration are received in a computer storage medium. A data de-duplication model is then applied to a set of data and to the data de-duplication parameters, and a hardware reliability model is applied to the hardware configuration. Then a set (at least one) of discrete events is simulated based on the data de-duplication model as applied to the set of data and the data de-duplication parameters, and the hardware reliability model as applied to the hardware configuration. Based on the simulation, a set of data reliability and availability estimations/estimates can be generated (e.g., and outputted/provided).

摘要翻译： 本发明的实施例提供一种利用离散事件模拟来定量分析计算机存储环境中的建模的重复数据删除系统的可靠性的方法。 另外，本文描述的方法可以在存在主异构存在的异构存储系统上的具有异构数据的系统上进行这种分析，并且由于重复数据删除而产生其次要效应。 在典型的实施例中，在计算机存储介质中接收数据重复数据删除参数和硬件配置。 然后将数据重复数据删除模型应用于一组数据和重复数据删除参数，并将硬件可靠性模型应用于硬件配置。 然后基于应用于数据集和数据重复数据删除参数的数据重复数据删除模型以及应用于硬件配置的硬件可靠性模型来模拟一组（至少一个）离散事件。 基于该仿真，可以生成（例如和输出/提供）一组数据可靠性和可用性估计/估计。

公开(公告)号：US20120109907A1

公开(公告)日：2012-05-03

申请号：US12916524

申请日：2010-10-30

申请人： Nagapramod S. Mandagere ,  David A. Pease ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Nagapramod S. Mandagere ,  David A. Pease ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F17/30

CPC分类号： G06F16/215

摘要： Embodiments of the invention relate to performing on-demand data deduplication for managing data and storage space. Redundant data in a system is detected. Availability of data storage space in the system is periodically evaluated. Performance parameters of the system are evaluated. Detected redundant data is selected based on the data storage availability and performance parameters of the system. If at least a portion of the selected redundant data is to be deduplicated is determined.

摘要翻译： 本发明的实施例涉及执行用于管理数据和存储空间的按需重复数据删除。 检测到系统中的冗余数据。 定期评估系统中数据存储空间的可用性。 对系统的性能参数进行评估。 根据系统的数据存储可用性和性能参数选择检测到的冗余数据。 如果确定了所选择的冗余数据的至少一部分被重复数据消除。

公开(公告)号：US20110295815A1

公开(公告)日：2011-12-01

申请号：US12788026

申请日：2010-05-26

申请人： Nagapramod S. Mandagere ,  Ramani R. Routray ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Nagapramod S. Mandagere ,  Ramani R. Routray ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F17/30

CPC分类号： G06F11/1004

摘要： Embodiments of the invention relate to testing a storage system point-in-time copy of data for consistency. An aspect of the invention includes receiving system and application event information from systems and applications associated with point-in-time copies of data. The system and application event information is associated with each of point-in-time copies of data. At least one point-in-time copy of data is selected for testing. The system and application event information is compared with inconsistency classes to determine tests for testing the point-in-time copy of data. The point-in-time copy of data is tested.

摘要翻译： 本发明的实施例涉及测试用于一致性的数据的存储系统时间点副本。 本发明的一个方面包括从与数据的时间点副本相关联的系统和应用接收系统和应用事件信息。 系统和应用事件信息与数据的每个时间点副本相关联。 选择至少一个时间点的数据副本进行测试。 将系统和应用程序事件信息与不一致类进行比较，以确定测试数据的时间点副本。 测试数据的时间点副本。

公开(公告)号：US08818962B2

公开(公告)日：2014-08-26

申请号：US12788026

申请日：2010-05-26

申请人： Nagapramod S. Mandagere ,  Ramani R. Routray ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Nagapramod S. Mandagere ,  Ramani R. Routray ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F17/00 ,  G06F17/30

CPC分类号： G06F11/1004

摘要： Embodiments of the invention relate to testing a storage system point-in-time copy of data for consistency. An aspect of the invention includes receiving system and application event information from systems and applications associated with point-in-time copies of data. The system and application event information is associated with each of point-in-time copies of data. At least one point-in-time copy of data is selected for testing. The system and application event information is compared with inconsistency classes to determine tests for testing the point-in-time copy of data. The point-in-time copy of data is tested.

摘要翻译： 本发明的实施例涉及测试用于一致性的数据的存储系统时间点副本。 本发明的一个方面包括从与数据的时间点副本相关联的系统和应用接收系统和应用事件信息。 系统和应用事件信息与数据的每个时间点副本相关联。 选择至少一个时间点的数据副本进行测试。 将系统和应用程序事件信息与不一致类进行比较，以确定测试数据的时间点副本。 测试数据的时间点副本。

公开(公告)号：US20130110793A1

公开(公告)日：2013-05-02

申请号：US13286490

申请日：2011-11-01

申请人： Kavita Chavda ,  Eric W. Davis Rozier ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Kavita Chavda ,  Eric W. Davis Rozier ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F7/00

CPC分类号： G06F7/00 ,  G06F17/30159

摘要： Embodiments of the present invention provide an approach that utilizes discrete event simulation to quantitatively analyze the reliability of a modeled de-duplication system in a computer storage environment. In addition, the approach described herein can perform such an analysis on systems having heterogeneous data stored on heterogeneous storage systems in the presence of primary faults and their secondary effects due to de-duplication. In a typical embodiment, data de-duplication parameters and a hardware configuration are received in a computer storage medium. A data de-duplication model is then applied to a set of data and to the data de-duplication parameters, and a hardware reliability model is applied to the hardware configuration. Then a set (at least one) of discrete events is simulated based on the data de-duplication model as applied to the set of data and the data de-duplication parameters, and the hardware reliability model as applied to the hardware configuration. Based on the simulation, a set of data reliability and availability estimations/estimates can be generated (e.g., and outputted/provided).

摘要翻译： 本发明的实施例提供一种利用离散事件模拟来定量分析计算机存储环境中的建模的重复数据删除系统的可靠性的方法。 另外，本文描述的方法可以在存在主异构存在的异构存储系统上的具有异构数据的系统上进行这种分析，并且由于重复数据删除而产生其次要效应。 在典型的实施例中，在计算机存储介质中接收数据重复数据删除参数和硬件配置。 然后将数据重复数据删除模型应用于一组数据和重复数据删除参数，并将硬件可靠性模型应用于硬件配置。 然后基于应用于数据集和数据重复数据删除参数的数据重复数据删除模型以及应用于硬件配置的硬件可靠性模型来模拟一组（至少一个）离散事件。 基于该仿真，可以生成（例如和输出/提供）一组数据可靠性和可用性估计/估计。

公开(公告)号：US20130006943A1

公开(公告)日：2013-01-03

申请号：US13172930

申请日：2011-06-30

申请人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F17/30

CPC分类号： G06F11/1458 ,  G06F11/1448 ,  G06F11/1464

摘要： Embodiments of the present invention provide a hybrid (e.g., local and remote) approach for data backup in a networked computing environment (e.g., a cloud computing environment). In a typical embodiment, a set of storage configuration parameters corresponding to a set of data to be backed up is received and stored in a computer data structure. The set of storage configuration parameters can comprise at least one of the following: a recovery time objective (RTO), a recovery point objective (RPO), and a desired type of protection for the set of data. Regardless, the set of data is compared to previously stored data to identify at least one of the following: portions of the set of data that have commonality with the previously stored data; and portions of the set of data that are unique to the set of data (i.e., not in common with any of the previously stored data). The above-described process is referred to herein as “de-duplication”. A storage solution is then determined based on the set of storage configuration parameters. In general, the storage solution identifies at least one local storage resource and at least one remote storage resource (e.g., a cloud storage resource) for backing up the portions of the set of data that are unique to the set of data. Once the storage solution has been determined, the unique portions of the set of data will be stored in accordance therewith.

摘要翻译： 本发明的实施例提供了一种用于网络计算环境（例如，云计算环境）中的数据备份的混合（例如，本地和远程）方法。 在典型的实施例中，接收一组对应于要备份的数据的存储配置参数并将其存储在计算机数据结构中。 该组存储配置参数可以包括以下中的至少一个：恢复时间目标（RTO），恢复点目标（RPO）以及该组数据的期望类型的保护。 无论如何，将该组数据与先前存储的数据进行比较，以识别以下中的至少一个：与先前存储的数据具有共同性的数据集合中的部分; 以及数据集合中对数据集合唯一的部分（即，与先前存储的任何数据不共同的）。 上述过程在本文中被称为重复数据删除。 然后基于存储配置参数的集合来确定存储解决方案。 通常，存储解决方案识别至少一个本地存储资源和至少一个远程存储资源（例如，云存储资源），用于备份该组数据的唯一的数据组的部分。 一旦确定了存储解决方案，则将根据其存储该组数据的唯一部分。

公开(公告)号：US20120330895A1

公开(公告)日：2012-12-27

申请号：US13168066

申请日：2011-06-24

申请人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F17/30

CPC分类号： G06F17/30581 ,  G06F17/30017 ,  G06F17/30283 ,  G06F17/30578

摘要： Embodiments of the present invention provide an approach for providing non-disruptive transitioning of application replication configurations and proactive analysis of possible error scenarios. Specifically, under embodiments of the present invention, a common integration model (CIM)-compatible representation of a system replication plan is provided in a computer data structure. Based on the representation, a hierarchical tree data structure having a set of nodes is created. A set of system configuration updates pertaining to the set of nodes are then classified (e.g., based upon the type of configuration update). Once the set of nodes has been classified, the set of nodes may then be analyzed to determine if any nodes of the set are isomorphic. If so, the plan can be modified accordingly. In any event, the replication plan (or modified replication plan) may then be implemented.

摘要翻译： 本发明的实施例提供了一种用于提供应用复制配置的非中断转换和可能的错误情景的主动分析的方法。 具体地，在本发明的实施例中，在计算机数据结构中提供了系统复制计划的公共集成模型（CIM）兼容表示。 基于该表示，创建具有一组节点的分层树数据结构。 然后对与该组节点有关的一组系统配置更新进行分类（例如，基于配置更新的类型）。 一旦节点集合被分类，则可以分析节点集合以确定集合中的任何节点是否是同构的。 如果是这样，可以相应地修改计划。 无论如何，可以实现复制计划（或修改的复制计划）。

公开(公告)号：US08495018B2

公开(公告)日：2013-07-23

申请号：US13168066

申请日：2011-06-24

申请人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

发明人： Kavita Chavda ,  Nagapramod S. Mandagere ,  Sandeep M. Uttamchandani ,  Pin Zhou

IPC分类号： G06F17/00 ,  G06F17/30

CPC分类号： G06F17/30581 ,  G06F17/30017 ,  G06F17/30283 ,  G06F17/30578

摘要： Embodiments of the present invention provide an approach for providing non-disruptive transitioning of application replication configurations and proactive analysis of possible error scenarios. Specifically, under embodiments of the present invention, a common integration model (CIM)-compatible representation of a system replication plan is provided in a computer data structure. Based on the representation, a hierarchical tree data structure having a set of nodes is created. A set of system configuration updates pertaining to the set of nodes are then classified (e.g., based upon the type of configuration update). Once the set of nodes has been classified, the set of nodes may then be analyzed to determine if any nodes of the set are isomorphic. If so, the plan can be modified accordingly. In any event, the replication plan (or modified replication plan) may then be implemented.

摘要翻译： 本发明的实施例提供了一种用于提供应用复制配置的非中断转换和可能的错误情景的主动分析的方法。 具体地，在本发明的实施例中，在计算机数据结构中提供了系统复制计划的公共集成模型（CIM）兼容表示。 基于该表示，创建具有一组节点的分层树数据结构。 然后对与该组节点有关的一组系统配置更新进行分类（例如，基于配置更新的类型）。 一旦节点集合被分类，则可以分析节点集合以确定集合中的任何节点是否是同构的。 如果是这样，可以相应地修改计划。 无论如何，可以实现复制计划（或修改的复制计划）。