Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
* Require current version of dbmsbenchmarker * TPC-H: Allow to set a default script (for example Create Schema) * Configuration: Draft yugabytedb using own get_service_sut() * Configuration: DBMSBenchmarker does not have to wait before starting * Docs: Sketch of workflow (parallel generate/load and benchmark * Configuration: Draft kinetica using own get_service_sut() * Configuration: Draft kinetica using own create_monitoring() * Configuration: Draft kinetica using own set_metric_of_config() * Configuration: Jobs are synched by Redis, they do not have to wait for fixed period of time * Configuration: Upload YCSB results to cluster * Cluster: get_service_endpoints() * Cluster: prometheus_config contains endpoints in new format * Configuration: Catch if container does not have ENV * Cluster: prometheus_config contains endpoints in new format * YCSB: Benchmark might contain INSERT and SCAN * Configuration: Upload YCSB results to cluster * Configuration: connections.config should contain all connections processed so far * YCSB: Benchmark ignore GC * Configuration: Kinetica monitoring tests * Cluster: get_service_endpoints() bug * HammerDB: Upload results to evaluation server * Cluster: endpoints_cluster for monitoring * Metrics: Option container_name * Metrics: Fetch loader metrics first try * Metrics: Fetch loader metrics by benchmarker section * Metrics: Fetch benchmarker metrics by benchmarker section * Metrics: Scraping and timeout 10s * AWS: Set context * Experiments: Only start monitoring if benchmarks left * Experiments: Store pod log, remove pod, then delete job * YCSB: Evaluator contains number of operations and works for workload E * YCSB: Averaging percentiles of latencies * Cluster: If monitor_cluster_active, do not deploy additional cAdvicors per SUT * Metrics: Fetch datagenerator metrics * Configuration: Catch if get_host_diskspace_used() returns weird format * HammerDB: Averaging metrics * Cluster: Job status checks number of completions * Configuration: Do not restart loading when sut pod got lost * Cluster: Job status checks number of completions >= * Configuration: Do not restart loading when sut pod got lost * Cluster: Job status checks number of completions not None * Cluster: Do not dump job status * Benchbase: Duration of pod is part of loading result * Evaluation: Know natural sort * Experiment: Option to change parameters for each run of a benchmarker instance * Experiment: Store pod logs before job has finished * Experiment: Store logs of job - message in debug * Experiment: Store logs of job - up to 10 retries * Configuration: Disk size in kilobytes * Experiment: Deploy daemonset * Cluster: Retry if timeout at execute_command_in_pod() * Removed IPython dependency
- Loading branch information