A server system receives an electronic document from a client system to be stored at a content management system. The server system generates multiple copies of the electronic document that are stored in parallel at respective multiple instances of the content management system. The server system receives a request to retrieve a copy of the electronic document. In response to the request, the server system checks, by a load balancer, whether the respective multiple instances of the content management system are operational to retrieve the copy of the electronic document. The server system then performs load balancing by the load balancer in accordance with a round-robin process to select a particular instance and retrieves the copy of the electronic document from the particular instance of the content management system.

Some embodiments provide a novel method for load balancing data messages that are sent by a source compute node (SCN) to one or more different groups of destination compute nodes (DCNs). In some embodiments, the method deploys a load balancer in the source compute node's egress datapath. This load balancer receives each data message sent from the source compute node, and determines whether the data message is addressed to one of the DCN groups for which the load balancer spreads the data traffic to balance the load across (e.g., data traffic directed to) the DCNs in the group. When the received data message is not addressed to one of the load balanced DCN groups, the load balancer forwards the received data message to its addressed destination. On the other hand, when the received data message is addressed to one of load balancer's DCN groups, the load balancer identifies a DCN in the addressed DCN group that should receive the data message, and directs the data message to the identified DCN. To direct the data message to the identified DCN, the load balancer in some embodiments changes the destination address (e.g., the destination IP address, destination port, destination MAC address, etc.) in the data message from the address of the identified DCN group to the address (e.g., the destination IP address) of the identified DCN.

The embodiments provide systems and methods for efficiently and accurately differentiating requests directed to uncacheable content from requests directed to cacheable content based on identifiers from the requests. The differentiation occurs without analysis or retrieval of the content being requested. Some embodiments hash identifiers of prior requests that resulted in uncacheable content being served in order to set indices within a bloom filter. The bloom filter then tracks prior uncacheable requests without storing each of the identifiers so that subsequent requests for uncacheable requests can be easily identified based on a hash of the request identifier and set indices of the bloom filter. Some embodiments produce a predictive model identifying uncacheable content requests by tracking various characteristics found in identifiers of prior requests that resulted in uncacheable content being served. Subsequent requests with identifiers having similar characteristics to those of the predictive model can then be differentiated.

Specialized, service optimized virtual machines are assigned to handle specific types of Internet of Things (IoT) devices. An IoT context mapping policy engine within the context of a virtualized network function manages IoT context mapping policy functions in load balancers. The IoT context mapping policy functions select service optimized virtual machines based on IoT device IDs, and assign those virtual machines to handle the devices. The IoT context mapping policy functions provide load data to the IoT context mapping policy engine. Based on the load data, the IoT context mapping policy engine maintains appropriate scaling by creating or tearing down instances of the virtual machines.

A system and method in accordance with example embodiments may include systems and methods for a cloud-based analytics platform. The cloud-based analytics platform may allow the manual and automatic uploading to and/or downloading from a cloud server. The platform may include single sign-on (SSO) capabilities such that a user may have one set of credentials to access data from the cloud-based analytics and/or data stored locally. The platform may include data validation and processing in order to provide real-time feedback on uploads based on file type, file size, access rights, extracted data, and transformed data.

Provided are a method and system for achieving high availability of service under a high-load scene in a distributed system. The method includes constructing a node selection model at a master node of a distributed cluster; constructing a request selection model in each slave node; wherein the request selection model includes weights of designated requests and trade-off parameters set for requests each having a weight greater than a set value; when the distributed system enters the high-load scene, reading, by the master node, the node selection model, and sequentially selecting the served slave nodes according to a time slice round robin policy; and reading, by each slave node, the request selection model, and sequentially returning data of each request according to the trade-off parameters of each request.

Provided are a method and system for achieving high availability of service under a high-load scene in a distributed system. The method includes constructing a node selection model at a master node of a distributed cluster; constructing a request selection model in each slave node; wherein the request selection model includes weights of designated requests and trade-off parameters set for requests each having a weight greater than a set value; when the distributed system enters the high-load scene, reading, by the master node, the node selection model, and sequentially selecting the served slave nodes according to a time slice round robin policy; and reading, by each slave node, the request selection model, and sequentially returning data of each request according to the trade-off parameters of each request.

Some embodiments provide a method of performing load balancing for a group of machines that are distributed across several physical sites. The method of some embodiments iteratively computes (1) first and second sets of load values respectively for first and second sets of machines that are respectively located at first and second physical sites, and (2) uses the computed first and second sets of load values to distribute received data messages that the group of machines needs to process, among the machines in the first and second physical sites. The iterative computations entail repeated calculations of first and second sets of weight values that are respectively used to combine first and second load metric values for the first and second sets of machines to repeatedly produce the first and second sets of load values for the first and second sets of machines. The repeated calculation of the weight values automatedly and dynamically adjusts the load prediction at each site without user adjustment of these weight values. As it is difficult for a user to gauge the effect of each load metric on the overall load, some embodiments use machine learned technique to automatedly adjust these weight values.

Some embodiments provide a method of performing load balancing for a group of machines that are distributed across several physical sites. The method of some embodiments iteratively computes (1) first and second sets of load values respectively for first and second sets of machines that are respectively located at first and second physical sites, and (2) uses the computed first and second sets of load values to distribute received data messages that the group of machines needs to process, among the machines in the first and second physical sites. The iterative computations entail repeated calculations of first and second sets of weight values that are respectively used to combine first and second load metric values for the first and second sets of machines to repeatedly produce the first and second sets of load values for the first and second sets of machines. The repeated calculation of the weight values automatedly and dynamically adjusts the load prediction at each site without user adjustment of these weight values. As it is difficult for a user to gauge the effect of each load metric on the overall load, some embodiments use machine learned technique to automatedly adjust these weight values.

System and methods are described for performing load balancing by continually collecting real-time metrics values from a plurality of endpoints in a cloud computing system, the real-time metrics values representing current performance measurements of processing by the endpoints, and using the collected real-time metrics values by a controller to continually determine a current weight value for each endpoint, the current weight value representing a probability that the endpoint will be selected to process a user request. The method includes receiving the user request for the cloud computing system to perform requested processing; selecting an endpoint of the cloud computing system to process the user request based at least in part on the current weight values of the endpoints; and sending the user request to the selected endpoint.