Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #8 from camunda-consulting/multiple-models
* execute benchmark with some chaosmesh experiments * run for 20 minutes and not 30 * bugfixes * set the vcpu request to limit-1 * change experimlent namespace * correct chaos brokers indentation * change the dashboard id * Adding new spreadsheet parameters * Fix message scenario (WiP) * Fix kill * Really fix kill * Add image sync for research prototypes * Stop on error * Make time ranges more precise * Reintroduce grafana target * Reduce to 15 min * Add sysout * Add TODO * Add tempfs volume for state data * Make process model dynamic * Only export scenarios with name and no time * Reduce to 10 min * Keep looking for new test cases * Fix directory iteration * Increase to 20 minutes * Remove dir * Back to 20 min * Add rocksdb.disableWal * Finally fix the kill ...until we switch to a job ;-) * Immediately persist test config * Reuse rebalance-leader from Helm profiles * share the 4 storage classes we used during last perf tests on GCP * Increase to 60 min * Exit after generation * Make run duration configurable * Update to job exporter 0.0.10 (with 550ms job delay) * Only sync non-standard images * Add TTL, exporter 0.0.7, generate sleep script * Add command-line instructions * Make sleep dynamic * Switch back to system properties * Add target `logs-benchmark` * Record Helm chart version * Deploy multiple models (WiP) * Deploy multiple BPMN and DMN models from a folder * Disable Identity * Disable Connectors * Disable Job Exporter * Disable ramdiks for data dir (tmpfs) * Reorder benchmark*.yaml for better diff & merge * Move models to separate dir; use official c8b image * Disable auto-commit to avoid leaking customer data * Add more maintenance targets * Copy models in order to preserve them in the run * Improve full stack profile (WiP) * Fix url-grafana target (and include for camunda.mk?) * Fix Grafana Dashboard URL * Ensure configmap entry names * Merge full-stack values into normal values * Make copy-models idempotent * Fix copy-models * Add install target to launch the runner * Add symlink to current run and `clean` target * Fix deploy job (for now by hard coding zeebe 8.2.10) * Fix removal of current/run * Check if image tag already exists and avoid pulling again * Fix debug symlink * Fix deployment without DMN files * Switch back to create process instance (w/o msg) * Improve `make clean` by running `make clean` of the current run * make root more dynamic so that rebalance is working * add numberOfThreads and maxJobsActive to benchmark * Turn on job streaming * 1.Changes to Elastic properties in sync with Helm chart 9.9.3 2.Configure number of shards for elastic exporter's indices 3.Added Raft flush delay time TODO: Eliminate ZEEBE_BROKER_EXPERIMENTAL_DISABLEEXPLICITRAFTFLUSH & use ZEEBE_BROKER_CLUSTER_RAFT_FLUSH_DELAYTIME to drive raft flush * Added ES number of shards as an input from the sheet. Added disk size for ES TODO: Eliminate ZEEBE_BROKER_EXPERIMENTAL_DISABLEEXPLICITRAFTFLUSH & use ZEEBE_BROKER_CLUSTER_RAFT_FLUSH_DELAYTIME to drive raft flush * revert file - comment max-jobs-active & numthreads * Jens wip (#9) * Update syntax for Elasticsearch subchart to CHC 9.X structure * Add Operate Config * Add Elasticsearch HeapSize * update es config, disable console&tasklist * Add Operate & ES Configs * fix newline --------- Merging changes into multiple-models. The Google spreadsheet is currently under review and will be revised again if required. ---------- Co-authored-by: jens lindner <[email protected]> Co-authored-by: Jothi Kiruthika <[email protected]> * update "${elasticSearch.shards}" -> "${elasticSearch.shardsPerIndex}" * Updated Makefile to include json payload file name from the excel test case * Update camunda-values.yaml - fix operate.importerThreadsCount * Revert camunda-values.yaml - changes to operate.importerThreadscount * Update application.yml - Added new input columns to the Excel template * Update README.md to refer the latest version of the benchmark test case template --------- Co-authored-by: Christophe <[email protected]> Co-authored-by: jens lindner <[email protected]> Co-authored-by: Jothi <[email protected]> Co-authored-by: Jothi Kiruthika <[email protected]>
- Loading branch information