Emc IP Holding Company LLC

Mellanox Technologies, Ltd.

Roman Spiegelman

Rivka Mayraz Matosevich

Doron Tal

Oren Duer

Eliav Bar-Ilan

Lior Zilpa

Evgeny Katz

German Goft

In one embodiment, a system includes a storage device controller including a first controller to read commands from a submission queue stored in a shared memory, provide the commands to a second controller, and write completion notices received from the second controller to a completion queue in the shared memory, and the second controller to receive the commands from the first controller, perform storage operations with a non-volatile memory responsively to receiving the commands, generate the completion notices responsively to performing the storage operations, provide the completion notices to the first controller, write recovery data about the commands and the completion notices to a persistent memory, and recover from a failure responsively to retrieving the recovery data from the persistent memory.

Transparent recovery of emulated storage device after a failure

Techniques are provided for managing replication journals in a distributed replication system. A data replication system assigns an associated replication journal volume to each replication component operating on a storage node. Each replication component is configured to handle a portion of a replication workload which is distributed among the replication components, and to write journal data in the associated replication journal volume of the replication component. The data replication system performs a recovery process in response to detecting a failed replication component. The recovery process includes designating at least one replication component as a recovery replication component, designating the associated replication journal volume of the failed replication component as a recovery journal volume, and assigning the recovery journal volume to the recovery replication component to enable the recovery replication component to recover journal data in the recovery journal volume.

Managing replication journal in a distributed replication system

Techniques are provided for managing system upgrades in a network computing environment. For example, a method includes managing an upgrade of a system within a cluster of systems which communicate in a network computing environment system to implement a distributed application. The system performs an upgrade operation to upgrade a current version of the system to a new software version number which comprises a new feature. The system performs a cluster version update process to communicate with other systems in the cluster to determine a cluster version number of the cluster. The system enables the new feature of the new software version if the new software version number is not greater than the determined cluster version number of the cluster.

Managing system upgrades in a network computing environment

A management server modifies a set of objects which is utilized by the management server to manage a network computing environment including server node. The management server updates a set of container objects in response to modifying the set of objects. Each container object includes attributes to specify a unique identifier of the container object, to specify an update generation number of the container object, and to specify a unique identifier of an object that is associated with the container object. An object synchronization process is performed using the updated set of container objects. The object synchronization process includes identifying each container object having update generation number which is greater than a last update generation number of the given server node, and sending the identified container objects to the given server node to update the set of local objects of the given server node based on the identified container objects.

Object synchronization of server nodes in a network computing environment

Techniques are provided for managing the utilization of shared memory resources among a plurality of asynchronous replication workloads. A data replication system maintains a respective replication journal for each replication workload in an allocated portion of system memory of a source server node. Each replication journal is divided into journal barrier intervals by respective individual barriers for transmission to a destination server node. The data replication system waits for a triggering event to occur for generating a common barrier, wherein the triggering event is configured to prevent exhaustion of the allocated portion of system memory for storing the replication journals. The common barrier is generated to close the replication journals of some or all of the replication workloads in advance of the respective individual barriers associated, and the replication journals closed by the common barrier are transmitted to the destination server node.

Growing community of inventors

Yokneam Illit, Israel

Roman Spiegelman