Skip to content

OutputMetrics

L. Wagner edited this page Sep 9, 2022 · 2 revisions

Output Metrics

MiSim supports the following output metrics. A csv table for each mertric is generated that contains data pairs of simulation time (column SimulationTime) and the assotiated value (column Value) .

# Metric File Desciption
1 System Load G[<LoadGeneratorName>]_[<TargetOperation>]_Load.csv Load accumulated over 1 STU
2 Failed Requests G[<LoadGeneratorName>]_[<TargetOperation>]_FailedRequests.csv Failed Requests accumulated over 1 STU
3 Successful Requests G[<LoadGeneratorName>]_[<TargetOperation>]_SuccessfulRequests.csv Successful Requests accumulated over 1 STU
4 Failed Requests Global GEN_ALL_FailedRequests.csv Accumulation of failed requests over all load generators binned by 1 STU.
5 Successful Requests Global GEN_ALL_SuccessfulRequests.csv Accumulation of successful requests over all Load generators binned by 1 STU.
6 Global Response Times R[All]ResponseTimes.csv All measured responsetimes, independent of service.
7 Response Times R[<EndpointName>]_ResponseTimes.csv Response times simualted for the specific endpoint.
8 Instance Count S[<ServiceName>]_InstanceCount.csv Number of Instances of the respective service.
9 Load Distribution S[<ServiceName>]_Load_Distribution.csv List of instances that were chosen as load balancing target.
10 Service CPU Utilization S[<ServiceName>]_CPUUtilization.csv Average relative utilization of all instances of the service. (Probed every 0.1 STU by default)
11 Autoscaling Decision AS[<servicename>]_Decision.csv Holds the autoscaler decision as "UP","HOLD" and "DOWN"
12 Instance Changes AS[<servicename>]_InstanceChange.csv Number of started/stopped instances for a scaling decision
13 Measured Utilization AS[<servicename>]_MeasuredUtilization.csv Periodically measured Utilization by the autoscaler
14 Requests in System I[<InstanceName>]_Requests_InSystem.csv Number of Requests currently handled by an Instance.
15 Not Computed Requests I[<InstanceName>]_Requests_NotComputed.csv Number of Requests that still have a remaining computational demand for the instances' CPU.
16 Requests Waiting For Dependencies I[<InstanceName>]_Requests_WaitingForDependencies.csv Number of Requests still waiting for dependenies (data collection) to complete.
17 Internal Request Sends I[<InstanceName>]_SendOff_Internal_Requests.csv Number of open Dependencies. (One request may produces multiple dependencie requests.)
18 Instance State I[<InstanceName>]_State.csv State of the instance (Running, Shutting Down, Shut Down, ...)
19 Active Processes C[_CPU]_ActiveProcesses.csv The count of requests that are currently using active CPU time. Essentially active Threads.
20 Relative Utilization C[_CPU]_RelativeUtiliization A measure of how long the CPU will take to complete its active work and clear its queue.
21 Total Processes C[_CPU]_TotalProcesses.csv Count of requests currently assigend to be handled by the CPU. (Queuelenght + Active Processes)
22 Utilization C[_CPU]_Utilization.csv Percentage utilization of the CPU changes in n/CoreCount sized steps.
23 Binned/Avg. Utilization C[_CPU]_UtilizationBinned.csv Running Average over the utilization with window size of 0.5 STU (not modifyable yet).
24 Network Latency NL_latency.csv List of calculated Network latencies. Disabled by default.
25 Circuit Breaker State CB[<CircuitBreakerName>]_[<CB_TargetService>].csv Tupel of current (State, #SuccessFullRequests,#FailedRequests, FailureRate)