Abstract: Some embodiments provide a method for evaluating locations of applications in a multi-cloud network with applications located in different cloud datacenters of one or more cloud providers. The method receives data for flows collected from the cloud datacenters. The data for each flow indicates a source and destination of the flow and applications to which at least one of the source and destination of the flow belong. The method uses network topology data to identify a subset of the flows as egress flows, each of which is charged by a cloud provider at which the flow source is located. Based on data from the cloud providers, the method determines a cost for each of the egress flows and a cost associated with each of the applications. The method provides these costs to users of the network for the users to optimize locations of the applications in the different cloud datacenters.

Abstract: Some embodiments provide a method for using flow-based load balancing to select a service endpoint from multiple service endpoints in a datacenter of an enterprise network for providing one or more services to client devices in the enterprise network. The method receives, from a network modeling appliance that categorizes service endpoints based on network data, a first set of service endpoints that provide at least one particular service for which a client connection is to be scheduled. The method generates an intersecting set of service endpoints based on the received first set of service endpoints and a second set of service endpoints identified by a load balancer that performs load balancing operations for data message flows to and from the plurality of service endpoints. Based on the generated intersecting set of service endpoints, the method selects a particular service endpoint for scheduling the client connection.

Abstract: Example methods and systems for load balancing with service-tier awareness are described. One example may involve a computer system receiving, from a client system, a service request that requires processing by one of multiple server pools that are reachable via the computer system. The multiple server pools may be associated with respective multiple service tiers. The computer system may obtain identity information identifying a user associated with the service request from the client system; and based on the identity information, mapping the service request to a particular service tier from the multiple service tiers. Next, a particular server pool that is associated with the particular service tier may be identified from the multiple server pools. The service request may be steered towards a destination server for processing, the destination server being selected from the particular server pool associated with the particular service tier.

Abstract: Example methods and systems for packet flow monitoring are described. In one example, a first computer system may detect a flow of packets along a datapath between a source and a destination and determine source attribute information associated with the source and destination attribute information associated with the destination. The first computer system may perform attribute-to-identifier mapping by (a) mapping the source attribute information to a source identifier having a reduced size compared to the source attribute information, and/or (b) mapping the destination attribute information to a destination identifier having a reduced size compared to the destination attribute information.

Abstract: Some embodiments provide a novel method of tracking connections in a network. The method receives an identification of a first network endpoint and a second network endpoint. The method then determines that the first network endpoint cannot directly address a packet flow to the second network endpoint. The method identifies an address translation rule of a network device that translates an address of the second network endpoint into a translated address. The method then determines that the first network endpoint can directly address a packet flow to the translated address. The method then identifies a route from the first network endpoint to the second endpoint through the network device that translates the address and displays the route including an identifier of the network device.

Abstract: Some embodiments provide a novel method of tracking connections in a network. The method receives an identification of a first network endpoint and a second network endpoint. The method then determines that the first network endpoint cannot directly address a packet flow to the second network endpoint. The method identifies an address translation rule of a network device that translates an address of the second network endpoint into a translated address. The method then determines that the first network endpoint can directly address a packet flow to the translated address. The method then identifies a route from the first network endpoint to the second endpoint through the network device that translates the address and displays the route including an identifier of the network device.

Abstract: Techniques for reducing the startup latency of functions in a Functions-as-a-Service (FaaS) infrastructure are provided. In one set of embodiments, a function manager of the FaaS infrastructure can receive a request to invoke a function uploaded to the infrastructure and can retrieve information associated with the function. The retrieved information can include an indicator of whether instances of the function may be sticky (i.e., kept in host system primary memory after function execution is complete), and a list of zero or more host systems in the FaaS infrastructure that currently have an unused sticky instance of the function in their respective primary memories. If the indicator indicates that instances of the function may be sticky and if the list identifies at least one host system with an unused sticky instance of the function in its primary memory, the function manager can select the at least one host system for executing the function.

Abstract: The disclosure provides an approach for distribution of functions among data centers of a cloud system that provides function-as-a-service (FaaS). For example, the disclosure provides one or more function distributors configured to receive a request for loading or executing a function, automatically determine an appropriate data center to load or execute the function, and automatically load or execute the function on the determined data center. In certain embodiments, the function distributors are further configured to determine an appropriate data center to provide storage resources for the function and configure the function to utilize the storage resources of the determined data center.

Abstract: A computer network optimization methodology is disclosed. In a computer-implemented method, components of a computing environment are automatically monitored, and have a feature selection analysis performed thereon. Provided the feature selection analysis determines that features of the components are in frequent communication and generating network latency. Provided the feature selection analysis determines that features of the components are not well defined, a similarity analysis of the features is performed. Results of the feature selection methodology are generated, and the components involved in the network traffic latency are reassigned to migrate the latency.

Abstract: The disclosure provides an approach for service deployment. Embodiments include receiving an indication of user intent for deployment of one or more services in a network from a user that is not an administrator of the network, wherein the indication of the user intent comprises a domain specific language (DSL). Embodiments include parsing the indication of the user intent to determine one or more constraints for deploying the one or more services. Embodiments include receiving topology information for the network, wherein the topology information comprises associations between racks and machines in the network. Embodiments include receiving network performance information for the network. Embodiments include determining one or more deployment rules for the one or more services based on the one or more constraints, the topology information, and the network performance information. Embodiments include deploying the one or more services in the network based on the one or more deployment rules.

Abstract: The disclosure provides an approach for distribution of functions among data centers of a cloud system that provides function-as-a-service (FaaS). For example, the disclosure provides one or more function distributors configured to receive a request for loading or executing a function, automatically determine an appropriate data center to load or execute the function, and automatically load or execute the function on the determined data center. In certain embodiments, the function distributors are further configured to determine an appropriate data center to provide storage resources for the function and configure the function to utilize the storage resources of the determined data center.

Abstract: Some embodiments provide a novel method of tracking connections in a network. The method receives an identification of a first network endpoint and a second network endpoint. The method then determines that the first network endpoint cannot directly address a packet flow to the second network endpoint. The method identifies an address translation rule of a network device that translates an address of the second network endpoint into a translated address. The method then determines that the first network endpoint can directly address a packet flow to the translated address. The method then identifies a route from the first network endpoint to the second endpoint through the network device that translates the address and displays the route including an identifier of the network device.

Abstract: The disclosure provides an approach for service deployment. Embodiments include receiving an indication of user intent for deployment of one or more services in a network from a user that is not an administrator of the network, wherein the indication of the user intent comprises a domain specific language (DSL). Embodiments include parsing the indication of the user intent to determine one or more constraints for deploying the one or more services. Embodiments include receiving topology information for the network, wherein the topology information comprises associations between racks and machines in the network. Embodiments include receiving network performance information for the network. Embodiments include determining one or more deployment rules for the one or more services based on the one or more constraints, the topology information, and the network performance information. Embodiments include deploying the one or more services in the network based on the one or more deployment rules.

Abstract: The disclosure provides an approach for distribution of functions among data centers of a cloud system that provides function-as-a-service (FaaS). For example, the disclosure provides one or more function distributors configured to receive a request for loading or executing a function, automatically determine an appropriate data center to load or execute the function, and automatically load or execute the function on the determined data center. In certain embodiments, the function distributors are further configured to determine an appropriate data center to provide storage resources for the function and configure the function to utilize the storage resources of the determined data center.

Abstract: In one example, a management node may include a storage device to store network device information associated with a plurality of network devices and physical hosts in a datacenter. Example network device information may include at least one of routing information and media access control (MAC) address information. Further, the management node may include a processor operable with the storage device and memory coupled to the processor. In one example, the memory may include a network topology generation unit to determine adjacency between the plurality of network devices in the datacenter using the routing information and/or media access control (MAC) address information associated with the plurality of network devices and generate a network topology including a network path between a source endpoint and a destination endpoint using the adjacency between the plurality of network devices.

Abstract: An approach for intercepting and caching Domain Name System (DNS) related data and sharing the cached DNS related data among hypervisors is provided. In an embodiment, a method comprises: receiving a DNS query from a virtual machine, determining whether a DNS reply to the DNS query has been received from a DNS server and whether an Internet Protocol (IP) address has been stored in a local DNS cache; in response to determining that tire DNS reply to the DNS query has been received from the DNS server and the IP address has been stored in the local DNS cache; retrieving the IP address from the local DNS cache; and providing the IP address to the virtual machine.

Abstract: A computer network optimization methodology is disclosed. In a computer-implemented method, components of a computing environment are automatically monitored, and have a feature selection analysis performed thereon. Provided the feature selection analysis determines that features of the components are in frequent communication and generating network latency. Provided the feature selection analysis determines that features of the components are not well defined, a similarity analysis of the features is performed. Results of the feature selection methodology are generated, and the components involved in the network traffic latency are reassigned to migrate the latency.

Abstract: A component awareness and proximity detection methodology is disclosed. In a computer-implemented method, components of a computing environment are automatically monitored, and configuration information used to uniquely identify the components and their corresponding physical residence in the computing environment and feature selection and location analysis performed thereon. Provided the feature selection analysis determines that features of the components are well defined and identified, a classification of the features is performed. Based on the classification of features components in the computing environment are selectively located in identifiable hosts is performed.

Abstract: In one example, a management node may include a storage device to store network device information associated with a plurality of network devices and physical hosts in a datacenter. Example network device information may include at least one of routing information and media access control (MAC) address information. Further, the management node may include a processor operable with the storage device and memory coupled to the processor. In one example, the memory may include a network topology generation unit to determine adjacency between the plurality of network devices in the datacenter using the routing information and/or media access control (MAC) address information associated with the plurality of network devices and generate a network topology including a network path between a source endpoint and a destination endpoint using the adjacency between the plurality of network devices.

Abstract: Techniques for reducing the startup latency of functions in a Functions-as-a-Service (FaaS) infrastructure are provided. In one set of embodiments, a function manager of the FaaS infrastructure can receive a request to invoke a function uploaded to the infrastructure and can retrieve information associated with the function. The retrieved information can include an indicator of whether instances of the function may be sticky (i.e., kept in host system primary memory after function execution is complete), and a list of zero or more host systems in the FaaS infrastructure that currently have an unused sticky instance of the function in their respective primary memories. If the indicator indicates that instances of the function may be sticky and if the list identifies at least one host system with an unused sticky instance of the function in its primary memory, the function manager can select the at least one host system for executing the function.