As a system developer, I would like to be able to define the number of cluster nodes my system has during the benchmarking phase, e.g., to check the scalability of the system by increasing the number of nodes.
As a user who is comparing systems, I would like to be able to compare distributed systems with non-distributed systems in a fair way, i.e., by defining the number of nodes available for the system as 1.
Solution
In the best case, the user has a slider with which he can decide on the number of system nodes vs. benchmark nodes.
the option should only be available when it makes sense, i.e., the HOBBIT platform is deployed on more than 2 worker nodes.
The workers should have IDs (e.g., with predefined labels) so that the deployment of containers can decide where to deploy which container. Another option would be to have a lot of predefined groups, e.g., "HOBBIT-SYSTEM-3-3, HOBBIT-BENCHMARK-4-2" which would represent that a worker belongs to the system cluster if benchmark vs. system nodes are 3 vs. 3 and it belongs to the benchmark cluster if the ratio is 4 vs. 2.
User story
As a system developer, I would like to be able to define the number of cluster nodes my system has during the benchmarking phase, e.g., to check the scalability of the system by increasing the number of nodes.
As a user who is comparing systems, I would like to be able to compare distributed systems with non-distributed systems in a fair way, i.e., by defining the number of nodes available for the system as 1.
Solution
In the best case, the user has a slider with which he can decide on the number of system nodes vs. benchmark nodes.
Additional context Might exclude itself with #114