github.com/oam-dev/kubevela@v1.9.11/design/api/vela-controller-params-reference.md (about)

     1  # KubeVela Controller Parameters Reference
     2  
     3  |          parameter          |  type  |              default              |                           describe                           |
     4  | :-------------------------: | :----: | :-------------------------------: | :----------------------------------------------------------: |
     5  |         use-webhook         |  bool  |               false               |                   Enable Admission Webhook                   |
     6  |      webhook-cert-dir       | string | /k8s-webhook-server/serving-certs |               Admission webhook cert/key dir.                |
     7  |        webhook-port         |  int   |               9443                |               Admission webhook listen address               |
     8  |        metrics-addr         | string |               :8080               |          The address the metric endpoint binds to.           |
     9  |   enable-leader-election    |  bool  |               false               | Enable leader election for controller manager. Enabling this will ensure there is only one active controller manager. |
    10  |  leader-election-namespace  | string |                ""                 | Determines the namespace in which the leader election configmap will be created. |
    11  |        log-file-path        | string |                ""                 |                  The file to write logs to.                  |
    12  |      log-file-max-size      |  int   |               1024                | Defines the maximum size a log file can grow to, Unit is megabytes. |
    13  |          log-debug          |  bool  |               false               |          Enable debug logs for development purpose           |
    14  |       revision-limit        |  int   |                50                 | revision-limit is the maximum number of revisions that will be maintained. The default value is 50. |
    15  | application-revision-limit  |  int   |                10                 | application-revision-limit is the maximum number of application useless revisions that will be maintained, if the useless revisions exceed this number, older ones will be GCed first.The default value is 10. |
    16  |  definition-revision-limit  |  int   |                20                 | definition-revision-limit is the maximum number of component/trait definition useless revisions that will be maintained, if the useless revisions exceed this number, older ones will be GCed first.The default value is 20. |
    17  |    app-config-installed     |  bool  |               true                | app-config-installed indicates if applicationConfiguration CRD is installed |
    18  | autogen-workload-definition |  bool  |               true                | Automatic generated workloadDefinition which componentDefinition refers to |
    19  |         health-addr         | string |               :9440               |          The address the health endpoint binds to.           |
    20  |       apply-once-only       | string |               false               | For the purpose of some production environment that workload or trait should not be affected if no spec change, available options: on, off, force. |
    21  |       storage-driver        | string |               Local               |         Application file save to the storage driver          |
    22  | application-re-sync-period  |  time  |                5m                 | Re-sync period for application to re-sync, also known as the state-keep interval. |
    23  |      reconcile-timeout      |  time  |                3m                 |           The timeout for controller reconcile.              |
    24  | system-definition-namespace | string |            vela-system            |     define the namespace of the system-level definition      |
    25  |    concurrent-reconciles    |  int   |                 4                 | The concurrent reconcile number of the controller. You can increase the degree of concurrency if a large number of CPU cores are provided to the controller. |
    26  |        kube-api-qps         |  int   |                50                 | The qps for reconcile k8s clients. Increase it if you have high concurrency. A small number might restrict the requests to the api-server which may cause a long waiting queue when there are a large number of inflight requests. Try to avoid setting it too high since it will cause large burden on apiserver. |
    27  |       kube-api-burst        |  int   |                100                | The burst for reconcile k8s clients. The usage of this parameter is similar to kube-api-qps. Setting it to be larger than kube-api-qps allows accepting more requests temporarily. |
    28  |      depend-check-wait      |  time  |                30s                | depend-check-wait is the time to wait for ApplicationConfiguration's dependent-resource ready. |
    29  |        oam-spec-var         | string |               v0.3                |         the oam spec version controller want to set-up       |
    30  |         pprof-addr          | string |                ""                 | The address for pprof to use while profiling, empty means disable. |
    31  |        perf-enabled         |  bool  |               false               | Enable performance logging for controllers, disabled by default. |
    32  
    33  ### Recommended Parameters for Scenarios with Various Scale
    34  
    35  | Scale |  #Nodes  | #Applications |   #Pods  | concurrent-reconciles | kube-api-qps | kube-api-burst |  CPU  | Memory |
    36  | :---: | -------: | ------------: | -------: | --------------------: | :----------: | -------------: | ----: | -----: |
    37  | Small |  < 200   |   < 3,000     | < 18,000 |              2        |      300     |      500       |   0.5 |   1Gi  |
    38  | Medium | < 500   |   < 5,000     | < 30,000 |              4        |      500     |      800       |     1 |   2Gi  |
    39  | Large | < 1,000  |   < 12,000    | < 72,000 |              4        |      800     |     1,000      |     2 |   4Gi  |
    40  
    41  > For details, read KubeVela Performance Test Report