Skip to end of metadata
Go to start of metadata

Project Brief - What is Kontour ?

As the edge adoption increases, clear definitions of the Edge KPIs are required to match the application to an Edge infrastructure/BP with the required capability. Different edge applications may require different characteristics from the Edge site, for example low latency, very high throughput and so on. With growing number of edge sites catering to different needs and use-cases, there needs to be a categorisation of edge sites based on certain criteria. There may be certain other requirements like security, encryption etc. Project Kontour targets to define Edge KPIs for the plethora of edge applications. Refer list of KPIs

Use Case Details:

Use CaseRequirementsKPI (Example*)
Video camera transmitting security footage.
  • Ultra low latency
  • High upload network throughput.

ULTRA_LOW_LATENCY: Latency range should be 0-2 milliseconds.

HIGH_UPLOAD_THPUT: Upload throughput should be 10-100 Gbps

Retail advertising targeting ads and information

for retail organisations

  • High computing (for encryption)
  • Low upload network throughput.

HIGH_IPS: Instructions/second processing should be of the order of 2000-4000

HIGH_PIPS: Parallel instructions/second processing should be of the order of 4000-8000 

LOW_UPLOAD_THPUT: Upload throughput should be 100 Mbps - 1 Gbps

*The values listed here are only for illustrative purpose only.


  1. Identify and define KPIs for edge applications. These should also be measurable. When we say ultra low latency, the range for this latency should be defined and measurable(via tests). This will benefit in selecting multiple edge sites based for an application and lead to availability and cost savings.
  2. Identify and categorise edge sites based on their capabilities. Define minimum capabilities for a site to be called as an Edge Site. These capabilities could be – Services, APIs, Orchestration, Hardware specs (basic set) etc. ?
  3. Categorise edge sites - Identify categories of edge sites and define a criteria for categorisation.




Email addressPTL - Elected  (Y/N)
Arif KhanVoerEir ABarif29march@gmail.comY
Manik SidanaVoerEir
Kandan KathirvelAT&
Deepak KatariaAT&


August, 2020 Update: 

S.NoKPI NameCategoryDescriptionMeasurement Unit
1KT_KPI_NET_JITTERNetworkMeasure of jitter between two workloadsmilliseconds
2KT_KPI_NET_TCP_BWNetworkNetwork TCP Bandwidth between two workloadsGbps
3KT_KPI_NET_PING_LATNetworkPing Lateny between two workloadsmilliseconds
4KT_KPI_NET_UDP_LATNetworkUDP Lateny between two workloadsmilliseconds
5KT_KPI_NET_UDP_THPUTNetworkUDP throughput between two workloadsGbps
6KT_KPI_NET_UDP_PKT_LOSSNetworkPacket loss between two workloadsPackets lost per million packets
7KT_KPI_COMP_CPU_SCOREComputeCPU Performance as per index scoreCPU Index Score
8KT_KPI_COMP_IPSComputeMeasure CPU performance as instructions per secondsInstructions per second
9KT_KPI_COMP_MEM_LATComputeMaximum Memory latency of a workloadnanoseconds
10KT_KPI_COMP_MEM_BWComputeMemory bandwidth of a workloadMibyte/sec
11KT_KPI_STG_SEQ_RW_LATENCYStorageRead Write latency in the workload for sequential readwrite operation in 70:30 ratio.microseconds
12KT_KPI_STG_SEQ_READ_ONLY_IOPSStorageStorage IOPS of a workload for sequential read-only operationInput/Output Operations Per Second
13KT_KPI_STG_SEQ_WRITE_ONLY_IOPSStorageStorage IOPS of a workload for sequential write-only operationInput/Output Operations Per Second
14KT_KPI_STG_RAND_RW_LATENCYStorageRead Write latency in the workload for random readwrite operation in 70:30 ratio.microseconds
15KT_KPI_STG_RAND_READ_ONLY_IOPSStorageStorage IOPS of a workload for random read-only operationInput/Output Operations Per Second
16KT_KPI_STG_RAND_WRITE_ONLY_IOPSStorageStorage IOPS of a workload for random write-only operationInput/Output Operations Per Second
17KT_KPI_K8S_API_JOB_CREATE_READ_DELETEAPIKubernetes API performance for job creation, read and deleteSuccess Rate %
18KT_KPI_K8S_API_POD_CREATE_DELETEAPIKubernetes API performance for pod creation and deleteSuccess Rate %
19KT_KPI_K8S_API_NS_CREATE_DELETEAPIKubernetes API performance for namespace creation and deleteSuccess Rate %
20KT_KPI_K8S_API_REPCONT_CREATE_SCALE_DELETEAPIKubernetes API performance for replication controller creation, scale and deleteSuccess Rate %
21KT_KPI_K8S_API_REPSET_CREATE_SCALE_DELETEAPIKubernetes API performance for replica set creation, scale and deleteSuccess Rate %
22KT_KPI_K8S_API_STATEFULSET_CREATE_SCALE_DELETEAPIKubernetes API performance for statefulset creation, scale and deleteSuccess Rate %
23KT_KPI_K8S_API_DEP_CREATE_READ_DELETEAPIKubernetes API performance for deployment creation, read and deleteSuccess Rate %
24KT_KPI_K8S_API_NODEPORT_CREATE_CHECK_DELETEAPIKubernetes API performance for nodeport creation, check and deleteSuccess Rate %
25KT_KPI_K8S_SEC_CIS_COMPLIANCEAPIKubernetes security score measured via CIS benchmarkSecurity Score
26KT_KPI_K8S_CONF_E2E_SIG_AUTHAPIKubernetes conformance for sig-auth Compliance %age
27KT_KPI_K8S_CONF_E2E_SIG_STORAGEAPIKubernetes conformance for sig-storageCompliance %age
28KT_KPI_K8S_CONF_E2E_SIG_INSTRUMENTATIONAPIKubernetes conformance for sig-instrumentationCompliance %age
29KT_KPI_K8S_CONF_E2E_SIG_APPSAPIKubernetes conformance for sig-appsCompliance %age
30KT_KPI_K8S_CONF_E2E_SIG_CONFORMANCEAPIKubernetes conformance for ConformanceCompliance %age
31KT_KPI_K8S_CONF_E2E_SIG_NODE_CONFORMANCEAPIKubernetes conformance for NodeConformanceCompliance %age
32KT_KPI_K8S_CONF_E2E_SIG_API_MACHINERYAPIKubernetes conformance for sig-api-machineryCompliance %age
33KT_KPI_K8S_CONF_E2E_SIG_NETWORKAPIKubernetes conformance for sig-networrkCompliance %age
34KT_KPI_K8S_CONF_E2E_SIG_CLIAPIKubernetes conformance for sig-cliCompliance %age
35KT_KPI_K8S_CONF_E2E_SIG_SCHEDULINGAPIKubernetes conformance for sig-schedulingCompliance %age
36KT_KPI_SVC_NGINX_MAX_LATENCYServiceMaximum latency in NGINX in serving requestsmilliseconds
37KT_KPI_SVC_REDIS_MAX_THPUT_1MSServiceRedis Server Max Throughput for Latency Under 1msrequests/sec
38KT_KPI_SVC_MONGO_THPUTServiceMongo throughput for database operations per secondoperations/sec
39KT_KPI_K8S_NODE_HEALTHServiceCheck health of K8S nodes to be in ready stateNodes in Ready state
40KT_KPI_K8S_SYS_NS_HEALTHHealthCheck health of all pods in kube-system namespace All pods in running state
41KT_KPI_OS_API_NOVA_CRUDAPIOpenStack Nova API CRUD performanceSuccess Rate %
42KT_KPI_OS_API_NEUTRON_CRUDAPIOpenStack Neutron API CRUD performanceSuccess Rate %
43KT_KPI_OS_API_GLANCE_CRUDAPIOpenStack Glance API CRUD performanceSuccess Rate %
44KT_KPI_OS_API_CINDER_CRUDAPIOpenStack Cinder API CRUD performanceSuccess Rate %
45KT_KPI_OS_API_REFSTACKAPIOpenStack API compliance based on RefStack guidelinesCompliance %age
46KT_KPI_OS_API_TEMPEST_NOVAAPIOpenStack Nova API functional verificationCompliance %age
47KT_KPI_OS_API_TEMPEST_NEUTRONAPIOpenStack Neutron API functional verificationCompliance %age
48KT_KPI_OS_API_TEMPEST_GLANCEAPIOpenStack Glance API functional verificationCompliance %age
49KT_KPI_OS_API_TEMPEST_CINDERAPIOpenStack Cinder API functional verificationCompliance %age
50KT_ENC_RSA_SIGN_4096EncryptionMeasure of 4096 bit RSA signaturesSignatures/second
51KT_ENC_RSA_SIGN_2048EncryptionMeasure of 2048 bit RSA signaturesSignatures/second
52KT_ENC_RSA_SIGN_1024EncryptionMeasure of 1024 bit RSA signaturesSignatures/second
53KT_ENC_RSA_SIGN_512EncryptionMeasure of 512 bit RSA signaturesSignatures/second
54KT_AES128_CBC_THPUT_256EncryptionMeasure of throughput for AES CBC throughput for 256 bytes blockGb/second
55KT_AES128_CBC_THPUT_1024EncryptionMeasure of throughput for AES CBC throughput for 1024 bytes blockGb/second
56KT_AES128_CBC_THPUT_8192EncryptionMeasure of throughput for AES CBC throughput for 8192 bytes blockGb/second

September,2020 Updates:

Presenting Kontour in AiKaan webinar: