Abstract: Examples can include (1) identifying, on a network, a source node and a destination node, the source node including at least one source node virtual machine (“VM”) to be replicated as a destination node VM on the destination node, (2) performing a full synchronization by copying disks used by the source node VM in a current operational state to the destination node VM, (3) scheduling start times for multiple update synchronizations of changed data between the source node VM and the destination node VM, the start times being scheduled at different time intervals, wherein a first time interval is greater than a second time interval, and (4) performing, at a switch-over time, a shutdown of the source node VM and transmitting data changes that are pending on the disk to the destination node. Various corresponding systems, methods, and non-transitory computer-readable media are also disclosed.

Abstract: A system and method for backing up workloads for multiple tenants of a cloud computing system are disclosed. A method of backing up workloads for multiple tenants of a computing system, includes determining that a first tenant has elected to maintain a copy of backups in an off-site location and a second tenant has elected to not maintain a copy of backups in an off-site location, directing first backup data of the first tenant to a backup storage device within the computing system and to a backup storage device at the off-site location, and directing second backup data of the second tenant to the backup storage device within the computing system, and not to the backup storage device at the off-site location.

Abstract: Backups of workloads in a multi-tenant cloud computing environment are efficiently queued and scheduled. A method of backing up workloads for multiple tenants of a cloud computing system, includes the steps of determining a pool of workloads to be backed up during a time window, the workloads including workloads for a first tenant and workloads for a second tenant, placing a backup task for backing up each of the workloads in the pool in a scheduling queue and prioritizing the backup tasks according to backup parameters specified by the tenants, sizes of workload data to be backed up, and hardware constraints of a backup storage device executing the backups, and instructing the backup storage device to execute the backup tasks according to an order of the backup tasks in the scheduling queue.

Abstract: Techniques are provided for reducing the amount of data that to be transferred in a hybrid cloud system in order to spawn a VM at a private cloud computing system from a template library stored in a public cloud computing system. Instead of storing full virtual machine templates, template libraries at the public cloud computing system are “differential” libraries. Differential templates in these differential libraries only include data that is different as compared with either a base template or another differential template. Assuming a private cloud system already stores a base image, the private cloud system may spawn a VM based on a particular template simply by downloading the appropriate differential templates. This technique reduces the total amount of data that needs to be downloaded from the public cloud system in order to spawn a particular VM.

Abstract: A network management server allocates network addresses to a network gateway connected to a public network after receiving an allocation request that specifies a number of network addresses to allocate to the network gateway. The network management server determines a first and second subnets of the public network having, respectively, first and second pools of available network addresses capable of being allocated to devices on the public network, and allocates one or more available network addresses from the first and second pools to the network gateway.

Abstract: Multiple tenants within a hybrid cloud computing system may need IP addresses to communicate over a computer network external to the hybrid cloud system (such as the Internet). IP addresses are a scarce resource, and each address can only be assigned to a single tenant. With multiple tenants competing for IP addresses, many request collisions may occur if tenants request IP addresses in a naïve manner, such as by requesting the next available IP address numerically. When a collision occurs, a tenant must request a different IP address. Instead, tenants request random IP addresses within a particular subnet in a random manner, thereby reducing the number of collisions that occur and improving the latency associated with requesting an IP address.

Abstract: System and method for managing public internet protocol (IP) addresses for a group of virtual data centers utilize a virtual overlay network to route communications between the virtual data centers and a public network through a virtual edge router of one of the virtual data centers using a public IP address assigned to that virtual edge router.

Abstract: A system and method for backing up virtual machines are disclosed. To better use the backup resources, an optimal set of virtual machines needing backup during a time window is generated. The optimal set depends on a total time for backing up each virtual machine needing backup and a cost metric that indicates a cost of not backing up the virtual machine during the time window. The optimal set also meets various system constraints that reflect the backup resource limitations. Performing backups according to the optimal set limits the number of missed backups and the number of service level agreement violations, thereby improving the backup performance without costly additions to its infrastructure.

Abstract: A system and method for backing up workloads for multiple tenants of a cloud computing system are disclosed. A method of backing up workloads for multiple tenants of a computing system, includes determining that a first tenant has elected to maintain a copy of backups in an off-site location and a second tenant has elected to not maintain a copy of backups in an off-site location, directing first backup data of the first tenant to a backup storage device within the computing system and to a backup storage device at the off-site location, and directing second backup data of the second tenant to the backup storage device within the computing system, and not to the backup storage device at the off-site location.

Abstract: Backups of workloads in a multi-tenant cloud computing environment are efficiently queued and scheduled. A method of backing up workloads for multiple tenants of a cloud computing system, includes the steps of determining a pool of workloads to be backed up during a time window, the workloads including workloads for a first tenant and workloads for a second tenant, placing a backup task for backing up each of the workloads in the pool in a scheduling queue and prioritizing the backup tasks according to backup parameters specified by the tenants, sizes of workload data to be backed up, and hardware constraints of a backup storage device executing the backups, and instructing the backup storage device to execute the backup tasks according to an order of the backup tasks in the scheduling queue.

Abstract: Techniques are provided for reducing the amount of data that to be transferred in a hybridcloud system in order to spawn a VM at a private cloud computing system from a template library stored in a public cloud computing system. Instead of storing full virtual machine templates, template libraries at the public cloud computing system are “differential” libraries. Differential templates in these differential libraries only include data that is different as compared with either a base template or another differential template. Assuming a private cloud system already stores a base image, the private cloud system may spawn a VM based on a particular template simply by downloading the appropriate differential templates. This technique reduces the total amount of data that needs to be downloaded from the public cloud system in order to spawn a particular VM.

Abstract: Multiple tenants within a hybrid cloud computing system may need IP addresses to communicate over a computer network external to the hybrid cloud system (such as the Internet). IP addresses are a scarce resource, and each address can only be assigned to a single tenant. With multiple tenants competing for IP addresses, many request collisions may occur if tenants request IP addresses in a naive manner, such as by requesting the next available IP address numerically. When a collision occurs, a tenant must request a different IP address. Instead, tenants request random IP addresses within a particular subnet in a random manner, thereby reducing the number of collisions that occur and improving the latency associated with requesting an IP address.

Abstract: A network management server allocates network addresses to a network gateway connected to a public network after receiving an allocation request that specifies a number of network addresses to allocate to the network gateway. The network management server determines a first and second subnets of the public network ha mg respectively, first and second pools of available network addresses capable of being allocated to devices on the public network, and allocates one or more available network addresses from the first and second pools to the network gateway.