Arctos Labs have developed a technology that can optimize workload placement in hybrid / multi / edge compute scenarios.

With the advent of big data, AI and IoT the amount of data consumed by applications will drastically increase. This implies that we will see increasing issues of moving data back and forth from distant data centers. These issues relate to cost, latency as well as overload of transport networks. Such scenarios opens up a need to optimize the placement of application components.

Our studies show that there could be a 200% cost increase in placing application components in a non-optimal way.

Our placement optimization technology uses cost models for edge and central compute as well as for connectivity to calculate the cost for each possible distribution possible. The technology also uses application component constraints, such as latency to make sure those constraints are met. This is acheived by having probes embedded into the infrastructure and capture real-time metrics on connectivity.

Our technology will constantly monitor and optimize the placement in order to cater for any faults occuring in the infrastructure layer, and consequently trigger a fall-back placement to second best alternatives. This way our placement will secure distributed applications are always up and running.

Please contact Mats Eriksson, Arctos Labs for a demo or discussion

Read more at the folder below…..

hybrid_cloud_small Handout_Folder