Versions Compared

Old Version 350

changes.mady.by.user Morgan Richomme

Saved on

compared with

New Version Current

changes.mady.by.user Marek Szwałkiewicz

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

 

...

  • results not bad. Remaining test errors due to DCAE changes but a patch is  in gate - no results due to error below.
  • The Makupsafe issue (jinja2 upstream dependency) is also responsible for the istanbul k8s installation issue

...

  • deployment OK but all smoke tests failing due to Makupsafe

...

 

Agenda:

  • Staging pipeline migration details in progress:
    • We have confirmation that VEXXHOST will be our build target
    • We will be using k8s as a service offered by VEXXHOST
    • Ideally we won't be responsible for any infra operations/issues - if we use k8s as a service LFIT will handle support
    • Pipeline triggers will be done in Github Actions
    • We will send a self-contained payload (zipped charts and a test execution command?) as the k8s can't be accessed externally
    • Results will be published when the run ends


 

Agenda:

  • SDC tests fail in different ways between 1st and 2nd runs. Right now only logs from 2nd run are saved which makes debugging difficult. Lets try to:
    • remove retries in pythonsdk tests (maybe tests are not 100% reproducible?)
    • if the above does not work - try to hack some workaround
  • CR with testkube chart is waiting for review

...

  • testsuite/+/135640
    • tox yaml linting was changed to ignore helm directory files as helm is not a valid yaml

 

Agenda:

  • Message router errors are seen in Kiali
  • Introduction of testkube as an optional test component
    • We will use upstream - original testkube operator umbrella chart
    • Tests will be provided in one of the integration repos
    • Later we can include it in chained-ci process

 

Agenda:

  • Azure migration halted
  • Daily post-configure pipeline fails on Prometheus installation - Marek Szwałkiewicz take a look at version pinning
  • All daily tests green
  • What is up with weekly?



 

Agenda:

  • Status of Azure migration Marek Szwałkiewicz 
  • State of release preparations
  • UNH lab for gating status
  • Gating unblocked


 

Agenda:

  • Status of Azure migration Marek Szwałkiewicz 
  • State of release preparations
  • UNH lab for gating
  • Gating failures due to libssl version change?


 

Agenda:

  • Decide should we start migration out of Azure
    • Andreas and Marek vote YES we should invest time into migration to UNH
  • Errors we see in SDC-be logs regarding decryption of auth header.
    • SDC meeting was cancelled this week, maybe we can try to chat about it next week.
  • Daily network issues in DT lab, do we have Dailies elsewhere?
    • Issues are caused by DNS problems in DT lab
    • Dailies are run only on DT lab - that is an issue
  • List of stale repositories https://jira.onap.org/projects/INT/issues/INT-2195 
  • Figure out the availability and parameters of LFN openstack Marek Szwałkiewicz 

 

Agenda:

  • Gating and dailies are down due to Nexus on azure being dead. Lets talk about this.
    • Switch to main nexus and get docker credentials
      • docker account credentials are not available outside jenkins
    • Next steps:
      • Escalate Azure ownership to LFN
      • Notify Cedric/raise issue for him about this
      • Michał Jagiełło will chat with Paweł about maybe calling Eric?
    • UNH option
      • create k8s env manually
      • switch gating jobs
      • create nexus proxy
  • Errors we see in SDC-be logs regarding decryption of auth header.
    • Maybe its SDC-fe adding second header with the same name? If yes we need to fix it.
    • Add ticket about it to SDC project
  • PNF registration failure (probably) caused by database access during stability tests being very slow, its worth to test it with sidecar disabled on mariadb?

Changes:


 

Agenda:

  • Gating and dailies are down due to Nexus on azure being dead. Lets talk about this.
  • Errors we see in SDC-be logs regarding decryption of auth header.
    • ?
  • Database access during stability tests is very slow, its worth to test it with sidecar disabled on mariadb
    • Marek Szwałkiewicz will try to disable sidecar and run tests as soon as gating is working again

Changes:


 

Agenda:

  • Migration of chained-ci to LFN repository - problem with artifacts fetching blocks the transition
    • Implement solution #1,5 which is saving artifacts as zip in automated way inside git repo so we don't need to do it manually AND we don't have to be tied to gitlab-ci
  • Errors we see in SDC-be logs regarding decryption of auth header. Its probably caused by python-sdk.
    • Basically we call api with 2 Basic auth headers
    • update of the sdc-be image was done at the same time as the bug started appearing
    • Talk with Michał Jagiełło on Monday

Changes:

Archives:

...

...

  • Morgan Richomme  and Sylvain Desbureaux are stepping down
    • end of contributions: end of March 2022
    • end of Orange platform: end of Jakarta release (with minimal support)
    • Azure / NH won't be touched (gating will work, staging, ...) but resources to be identified to manage them...

 

  • Lab status
    • onap-k8s issue - there is an another issue after the dependency library new version release
    • Orange lab - some issues with connection during random tests
    • Jira
      serverONAP Jira
      columnIdsissuekey,summary,issuetype,created,updated,duedate,assignee,reporter,priority,status,resolution
      columnskey,summary,type,created,updated,due,assignee,reporter,priority,status,resolution
      serverId425b2b0a-557c-3c0c-b515-579789cceedb
      keyAAI-3450
      still blocks master gating
  • user-98b79 http://scancode.onap.eu/

 

 

...

...

Today morning Trivy was ran on ONAP daily master

Code Block
languagetext
Pods
onap-cli-757c76f546-zfd4m in vulnerable!
onap-message-router-0 in vulnerable!
onap-sdnrdb-coordinating-only-6f6d6b887b-kzr9x in vulnerable!
onap-sdnrdb-master-0 in vulnerable!
onap-sdnrdb-master-1 in vulnerable!
onap-sdnrdb-master-2 in vulnerable!
onap-vfc-huawei-vnfm-driver-57fddd8bc7-nqn7b in vulnerable!
onap-vnfsdk-56d9fb886f-sk4s4 in vulnerable!

Images:
nexus3.onap.org:10001/onap/cli:6.0.1 -> onap-cli-757c76f546-zfd4m
nexus3.onap.org:10001/onap/dmaap/dmaap-mr:1.3.0 -> onap-message-router-0
docker.nexus.azure.onap.eu/bitnami/elasticsearch:7.9.3 -> onap-sdnrdb-coordinating-only-6f6d6b887b-kzr9x, onap-sdnrdb-master-X
nexus3.onap.org:10001/onap/vfc/nfvo/svnfm/huawei:1.3.9 -> onap-vfc-huawei-vnfm-driver-57fddd8bc7-nqn7b
nexus3.onap.org:10001/onap/vnfsdk/refrepo:1.6.3 -> onap-vnfsdk-56d9fb886f-sk4s4

...

 

 

  • Nexus doesn't work - tests doesn't work....
  • When it worked we had several tests issues
    • needed to check if we don't use same kubeconfig on many tests pipelines
      • that can create conflicts because one test tries to create an instance of existing resource
  • Kubescape tests
    • could be useful but won't replace already existing tests
      • root pods check based just on the kubernetes resource definition (pod, deployment, statefulset etc.) and not check if the root user is disabled on the image
    • could be possible to add them into pipeline soon

...

  • there are currently 3 Azure labs
    • 2 gatings and 1 integration lab with ONAP
  • needs to ask TSC if it's possible to add one more to run daily tests - outside of any company infrastructure (currently Orange, Nokia, DT)
  • we could run daily on the 1 integration lab machine, but there is a risk we won't have a working lab if during daily execution ONAP creation fails

...

  • it's not required and will take a lot of time and resources, leave it as it is

...