Skip to end of metadata
Go to start of metadata

The purpose of this is to measure the load that a complete ONAP platform puts on a Kubernetes multi-node cluster orchestrated and deployed by a Rancher instance. Due to the nature of the deployment in an OOM-based ONAP platform which uses HELM for deployment, it is useful to have a standard set of initial resource requirements so that we know how much should be initially allocated for the Kubernetes node availability so as to not bring the cluster to a failure state due to resource availability concerns.

Rancher’s default Kubernetes template provides three components that have been used here to measure metrics. All of them run on pods within the Kubernetes cluster so the cluster itself can control it:

  • Heapster: This is a component that collects metrics and events from all the Kubernetes nodes. It is designed to be modular in a way that multiple backends are supported, such as InfluxDB (used here), Kakfa, ElasticSearch, Google Cloud Monitoring, etc.

  • InfluxDB: This is the backend component that Heapster uses to store the collected metric data (i.e. Heapster’s Sink). It is a time-series database that can be queried by many visualization GUIs, including Grafana and Kibana.

  • Grafana: This is the frontend component that queries InfluxDB to obtain the metrics collected by Heapster. It can contain many dashboards broken up into panels that can be customized to show all types of data provided by the backend data source.


In order to run these measurements, the following was applied:

  • An 11-node Kubernetes cluster (as currently being tested by the Integration team) is suggested to deploy a complete ONAP platform
    • Docker version 17.03.2
    • Kubernetes version 1.8.10

  • A single Rancher instance (v1.6.14)
    • Docker version 17.03.2

  • The VM flavor for each of these VMs is OpenStack standard flavor “m1.xlarge” which is equivalent to 8 vCPUs, 16 GB RAM, 160 GB HDD

  • ONAP "Beijing" Release (Complete, with all components enabled as part of the HELM chart deployment). Gerrit branch "BEIJING"


By using the components to measure resource usage (Heapster), store into a backend component (InfluxDB), and visualizing the data into a frontend (Grafana), the following relevant metrics were collected.


RESOURCE USAGE CHART

These metrics are from Kubernetes resources that do not stop after an amount of time (i.e. Kubernetes Jobs whose container exits as soon as the job request is completed), so focus is on Kubernetes Pods.


NOTE: All the numbers below are assuming that there is no additional actions on the ONAP applications, other than the resources running at an idle state.


  • For the complete ONAP instance usage:  this is measuring the mean value of overall kubernetes cluster usage after the sum of all the Kubernetes nodes resource usage reports. This takes into account the resource usage of a complete ONAP instance (with all its HELM Charts enabled and deployed).

RESOURCEvCPUsFILESYSTEMMEMORY (RAM)NETWORK

Complete ONAP Instance

(All 25 HELM Chart Components -

Beijing Release)

Min: 556 mcores

Max: 23.46 cores

Avg: 8.87 cores

Min: 2.84 GB

Max: 25.22 GB

Avg: 17.61 GB

Min: 8.5 GiB

Max: 130.1 GB

Avg: 118.7 GB

Min (Tx / Rx): 293 kBps / 282 kBps

Max (Tx / Rx): 2.09 MBps / 9.73 MBps

Avg (Tx / Rx): 371 kBps / 341 kBps


OVERALL CLUSTER - CPU USAGE


OVERALL CLUSTER - FILESYSTEM USAGE


OVERALL CLUSTER - MEMORY (RAM) USAGE


OVERALL CLUSTER - NETWORK USAGE


  • For measurement of aggregated pod usage (in other rows):

    • There is a TOTAL number of 165 Pods (157 in Running state / 8 in Completed state)


RESOURCEvCPUs (mcores)RAM MEMORY (GiB)NETWORK (kBps)

HELM AAF - Overall Pod Usage

  • TOTAL of 14 pods (13 in Running state / 1 in Completed state)
  • 3-node AAF quorum-client Cluster (3 K8S Replica Pods)

Min: 7

Max: 22

Avg: 11

Min: 10.23

Max: 10.68

Avg: 10.43

Min (Tx / Rx): 1.84 / 2.35

Max (Tx / Rx): 2.22 / 2.84

Avg (Tx / Rx): 2.00 / 2.52

HELM AAI - Overall Pod Usage

  • TOTAL of 16 pods (15 in Running state / 1 in Completed state)
  • 3-node AAI Cassandra Cluster (3 K8S Replica Pods)

Min: 1068

Max: 1635

Avg: 1191

Min: 10.52

Max: 11.39

Avg: 10.89

Min (Tx / Rx): 26.64 / 36.31

Max (Tx / Rx): 50.60 / 45.78

Avg (Tx / Rx): 39.79 / 40.69

HELM APPC - Overall Pod Usage

  • TOTAL of 4 pods (4 in Running state / 0 in Completed state)

Min: 14

Max: 25

Avg: 17

Min: 2.71

Max: 2.81

Avg: 2.74

Min (Tx / Rx): 0.249 / 0.321

Max (Tx / Rx): 0.335 / 0.394

Avg (Tx / Rx): 0.281 / 0.357

HELM CLAMP - Overall Pod Usage

  • TOTAL of 5 pods (5 in Running state / 0 in Completed state)

Min: 9

Max: 30

Avg: 13

Min: 4.26

Max: 4.36

Avg: 4.3

Min (Tx / Rx): 3.82 / 2.79

Max (Tx / Rx): 4.01 / 5.07

Avg (Tx / Rx): 3.92 / 4.98

HELM CLI - Overall Pod Usage

  • TOTAL of 1 pod (1 in Running state / 0 in Completed state)

Min: 0

Max: 0

Avg: 0

Min: 6.637 MiB

Max: 6.641 MiB

Avg: 6.639 MiB

Min (Tx / Rx): 0.027 / 0.053

Max (Tx / Rx): 0.045 / 0.064

Avg (Tx / Rx): 0.029 / 0.055

HELM CONSUL - Overall Pod Usage

  • TOTAL of 4 pods (4 in Running state / 0 in Completed state)
  • 3-node Consul Cluster (3 K8S Replica Pods)

Min: 408

Max: 514

Avg: 488

Min: 0.507

Max: 2.49

Avg: 1.14

Min (Tx / Rx): 44.33 / 758

Max (Tx / Rx): 51.03 / 856

Avg (Tx / Rx): 48.01 / 834

HELM DCAEGEN2 - Overall Pod Usage

  • TOTAL of 21 pods (20 in Running state / 1 in Completed state)
  • 11 Pods deployed by HELM templates, 10 Pods deployed by the Cloudify Manager Pod
  • 2-node DB Cluster (2 K8S Replica Pods)
  • 6-node Redis Cluster (6 K8S Replica Pods)
  • NOTE: Read metrics like (DCAE-HELM-METRICS) + (DCAE-CLOUDIFY-METRICS)

Min: (26) + (87) = 113

Max: (40) + (551) = 591

Avg: (30) + (251) = 281

Min: (2.06) + (4.70) = 6.76

Max: (2.11) + (6.38) = 8.49

Avg: (2.08) + (5.50) = 7.58

Min (Tx / Rx): (32.12) + (0.951) = 33.07 / (33.03) + (1.63) = 34.66

Max (Tx / Rx): (34.95) + (1.40) = 36.35 / (36.02) + (2.14) = 38.16

Avg (Tx / Rx): (33.50) + (1.09) = 34.59 / (34.49) + (1.80) = 36.29

HELM DMAAP - Overall Pod Usage

  • TOTAL of 6 pods (6 in Running state / 0 in Completed state)
  • 2-node dbc-pg Cluster (2 K8S Replica Pods)

Min: 126

Max: 178

Avg: 155

Min: 2.36

Max: 2.53

Avg: 2.44

Min (Tx / Rx): 163.30 / 180.25

Max (Tx / Rx): 195.81 / 212.87

Avg (Tx / Rx): 188.81 / 205.66

HELM ESR - Overall Pod Usage

  • TOTAL of 2 pods (2 in Running state / 0 in Completed state)

Min: 1

Max: 4

Avg: 2

Min: 545.6 MiB

Max: 570.9 MiB

Avg: 555 MiB

Min (Tx / Rx): 0.062 / 0.106

Max (Tx / Rx): 0.076 / 0.120

Avg (Tx / Rx): 0.069 / 0.109

HELM LOG - Overall Pod Usage

  • TOTAL of 13 pods (13 in Running state / 0 in Completed state)
  • 11-node logstash Cluster (11 K8S Replica Pods triggered by a DaemonSet)

Min: 232

Max: 26323

Avg: 3330

Min: 12.89

Max: 18.62

Avg: 16.30

Min (Tx / Rx): 6.09 / 31.38

Max (Tx / Rx): 180.82 / 97.80

Avg (Tx / Rx): 36.24 / 44.73

HELM MSB - Overall Pod Usage

  • TOTAL of 5 pods (5 in Running state / 0 in Completed state)

Min: 9

Max: 25

Avg: 14

Min: 657.2 MiB

Max: 713.6 MiB

Avg: 694 MiB

Min (Tx / Rx): 2.86 / 3.60

Max (Tx / Rx): 3.37 / 4.58

Avg (Tx / Rx): 3.08 / 3.94

HELM MULTICLOUD - Overall Pod Usage

  • TOTAL of 4 pods (4 in Running state / 0 in Completed state)

Min: 9

Max: 15

Avg: 13

Min: 1.14

Max: 1.21

Avg: 1.17

Min (Tx / Rx): 18.85 / 0.755

Max (Tx / Rx): 21.45 / 0.835

Avg (Tx / Rx): 20.13 / 0.795

HELM NBI - Overall Pod Usage

  • TOTAL of 3 pods (3 in Running state / 0 in Completed state)

Min: 7

Max: 10

Avg: 9

Min: 1.31

Max: 1.72

Avg: 1.57

Min (Tx / Rx): 0.828 / 0.869

Max (Tx / Rx): 0.849 / 0.904

Avg (Tx / Rx): 0.840 / 0.880

HELM OOF - Overall Pod Usage

  • TOTAL of 9 pods (9 in Running state / 0 in Completed state)

Min: 535

Max: 652

Avg: 578

Min: 10.85

Max: 11.20

Avg: 11.04

Min (Tx / Rx): 4.72 / 6.39

Max (Tx / Rx): 4.97 / 6.85

Avg (Tx / Rx): 4.88 / 6.68

HELM POLICY - Overall Pod Usage

  • TOTAL of 6 pods (6 in Running state / 0 in Completed state)

Min: 8

Max: 26

Avg: 14

Min: 6.74

Max: 7.33

Avg: 7.07

Min (Tx / Rx): 12 / 11.14

Max (Tx / Rx): 17.04 / 16.97

Avg (Tx / Rx): 16.11 / 15.15

HELM PORTAL - Overall Pod Usage

  • TOTAL of 7 pods (6 in Running state / 1 in Completed state)

Min: 24

Max: 209

Avg: 31

Min: 10.87

Max: 11.36

Avg: 11.17

Min (Tx / Rx): 3.80 / 4.84

Max (Tx / Rx): 25.50 / 17.10

Avg (Tx / Rx): 11.06 / 15.41

HELM ROBOT - Overall Pod Usage

  • TOTAL of 1 pod (1 in Running state / 0 in Completed state)

Min: 0

Max: 0

Avg: 0

Min: 5.13

Max: 5.13

Avg: 5.13

Min (Tx / Rx): 0.027 / 0.053

Max (Tx / Rx): 0.032 / 0.059

Avg (Tx / Rx): 0.029 / 0.054

HELM SDC - Overall Pod Usage

  • TOTAL of 10 pods (7 in Running state / 3 in Completed state)

Min: 268

Max: 440

Avg: 309

Min: 6.68

Max: 11.15

Avg: 8.88

Min (Tx / Rx): 41.87 / 44.65

Max (Tx / Rx): 86.53 / 91.36

Avg (Tx / Rx): 71.59 / 76.78

HELM SDNC - Overall Pod Usage

  • TOTAL of 7 pods (7 in Running state / 0 in Completed state)

Min: 1018

Max: 1032

Avg: 1023

Min: 2.81

Max: 2.78

Avg: 2.61

Min (Tx / Rx): 0.707 / 0.855

Max (Tx / Rx): 0.948 / 1.25

Avg (Tx / Rx): 0.407 / 1.16

HELM SNIRO-EMULATOR - Overall Pod Usage

  • TOTAL of 1 pods (1 in Running state / 0 in Completed state)

Min: 0

Max: 0.2

Avg: 0.001

Min: 525.99 MiB

Max: 526.08 MiB

Avg: 526.05 MiB

Min (Tx / Rx): 0.026 / 0.050

Max (Tx / Rx): 0.030 / 0.061

Avg (Tx / Rx): 0.028 / 0.055

HELM SO - Overall Pod Usage

  • TOTAL of 2 pods (2 in Running state / 0 in Completed state)

Min: 5

Max: 34

Avg: 6

Min: 2.63

Max: 3.77

Avg: 3.15

Min (Tx / Rx): 1.39 / 1.64

Max (Tx / Rx): 1.71 / 2.69

Avg (Tx / Rx): 1.52 / 1.79

HELM UUI - Overall Pod Usage

  • TOTAL of 2 pods (2 in Running state / 0 in Completed state)

Min: 934

Max: 1011

Avg: 991

Min: 1.35

Max: 1.51

Avg: 1.49

Min (Tx / Rx): 44.23 / 93.15

Max (Tx / Rx): 48.33 / 101.76

Avg (Tx / Rx): 47.23 / 99.56

HELM VFC - Overall Pod Usage

  • TOTAL of 17 pods (17 in Running state / 0 in Completed state)

Min: 395

Max: 462

Avg: 428

Min: 7.57

Max: 11.32

Avg: 9.56

Min (Tx / Rx): 0.898 / 1.66

Max (Tx / Rx): 1.04 / 1.83

Avg (Tx / Rx): 0.953 / 1.71

HELM VID - Overall Pod Usage

  • TOTAL of 3 pods (2 in Running state / 1 in Completed state)

Min: 3.2

Max: 9.8

Avg: 4.3

Min: 1.95

Max: 2.15

Avg: 2.05

Min (Tx / Rx): 0.708 / 0.552

Max (Tx / Rx): 0.846 / 0.648

Avg (Tx / Rx): 0.768 / 0.601

HELM VNFSDK - Overall Pod Usage

  • TOTAL of 2 pods (2 in Running state / 0 in Completed state)

Min: 0

Max: 0.2

Avg: 0.001

Min: 337.9 MiB

Max: 355.4 MiB

Avg: 348.9 MiB

Min (Tx / Rx): 0.013 / 0.026

Max (Tx / Rx): 0.016 / 0.034

Avg (Tx / Rx): 0.014 / 0.027




3 Comments

  1. Can you indicate clustering if it was on for any of the components like APPC/SDNC ?

    How would we use the per Project Data to "tune" the helm charts so helm does a better initial distribution ?


    Maybe add a kubectl -n onap get pods listing ?

    1. Hi Brian, sorry I have been working on filling out the numbers and took your advice into account. As you can see in the green table, I am pointing out where there is more than one replica wherever it's doing that. All these definitions are default in the Beijing (ONAP-2.0.0) branch release.


      I can also add the complete list of the pods (165 in total) but they are described in the green table next to their respective HELM chart layout.


      In order to tune the HELM charts, I would initially say that the HELM charts that have the heaviest loads get priority over deployment than the lightweight charts. Regardless of this fact, an "umbrella" chart like local/onap (holding all ONAP subcharts) will create k8s resources coming from the dependent subcharts, and they are all added up into a single set of k8s resources. Then, this aggregated set of k8s resources are sorted by type (the ordered list by which the k8s resources get deployed is here: https://github.com/helm/helm/blob/master/pkg/tiller/kind_sorter.go#L26).