2023-08-31

  • CEIP-4139: capability testing
    • investigate ways to connect to cluster to no avail
    • convo w Garrett
    • abortive chat w Ashish
    • write RFP

2023-08-30

  • kyverno
    • purge removes role and role_binding
    • secrets
      • should not be needed in kube-system.
      • are needed and self-signed we think elsewhere
      • Example in alpha: core-engineering-test-cluster-alpha
        • kube-system kyverno-svc.kube-system.svc.kyverno-tls-ca Opaque 1 11d
        • kube-system kyverno-svc.kube-system.svc.kyverno-tls-pair kubernetes.io/tls 2 11d creationTimestamp: “2023-08-18T17:11:22Z”
        • 3h after deletion

2023-08-29

  • CEIP-4139: capability testing:
    • discussion with Garrett (81mins):
      • capability testing has nothing to do with crtxctl
      • general agreement about utility / promising nature of Robot Framework
      • explained that crtxctl is for multi-cluster access and repl.
        • failed to demo well and neglected to mention RFP exists
      • likely he will now attempt alternative
      • agreed to regroup with RF working
    • decided fastest way to demo is calico not fluentbit
      • got basics into shape for ‘Access resource in other namespace when not isolated’
      • next: convert pod exec into real curl command

2023-08-28: public holiday

2023-08-25

  • intro types into crtxctl service config
  • CEIP-4355: cleanup typha-ca

2023-08-24

  • intro types into crtxctl service config
    • TIL: pydantic_settings BaseSettings
  • chat with Khush and Daniel about Adviser

2023-08-21

  • dupe CRDs
  • pixie diffs
  • support Khush RFP and command tests
  • CEIP-4344: cleanup typha-ca

2023-08-18

  • troubleshoot updates

2023-08-17

  • 1-2-1

    • pitch for Chris, DeAnna for how to engage PMs
  • CEIP-2614: labelling std review -> test resources

  • CEIP-4241: argo prod migration

    • CEIP-4291: hm-core-platform
  • Look at troubleshoot update following closure of: https://github.com/replicatedhq/troubleshoot/issues/1157

    kubectl krew update (turn off zscaler)
    ...
    Upgrades available for installed plugins:
      * krew v0.4.3 -> v0.4.4
      * preflight v0.57.0 -> v0.70.3
      * support-bundle v0.57.0 -> v0.70.3
    kubectl krew upgrade preflight
    kubectl krew upgrade support-bundle
    

login to argo

login argocd.cortex.elsevier.systems –sso set cluster interested in export C=hm-core-platform-cluster-prod-prod

create list of apps for cluster

argocd app list | grep hm-core-platform-cluster-prod-prod > hm-core-platform-cluster-prod-prod.txt

ascertain how many components (usually 22 or 21)

cat hm-core-platform-cluster-prod-prod.txt | cut -d ’ ’ -f -1 #21

create diff for each app

for c in cat hm-core-platform-cluster-prod-prod.txt | cut -d ' ' -f -1 ; do argocd app diff $c > ${c///-}.diff.txt; done ;

aggregate all diffs excluding argocd labels

cat *diff.txt | grep -v argocd > non-argo-diff.txt

eyeball the diffs

cat non-argo-diff.txt | more

only comments, checksums and crt diff, so apply

for c in cat hm-core-platform-cluster-prod-prod.txt | cut -d ' ' -f -1 ; do argocd app sync $c; done ;

verify

for c in cat hm-core-platform-cluster-prod-prod.txt | cut -d ' ' -f -1 ; do argocd app get $c | grep OutOfSync; done ;

2023-08-16

  • training
    • beyond video: operating model and workload suitability
  • half day

2023-08-15

  • training
    • benefits video

2023-08-14

  • CEIP-4042: move crtxctl to own repo
    • tidy up, write up follow on tasks
  • training
  • Retro
  • beta rollout of Argo
  • investigate cruft in kong-nonprod: CEIP-4328

2023-08-08-11

  • CEIP-4042: move crtxctl to own repo
    • with FG help implement RELEASE workflow compatible with terraform approach because already well known
    • bumpversion and publish to artifactory workflow
    • test and coverage workflow
    • sonarqube workflow
    • fix: single source version
    • fix broken relative paths
    • fix remote access to sbom

2023-08-07 - Vacation

2023-08-04

  • still fighting RDP, apparently did logging in wrong region.

  • purge q alternative:

    aws sqs delete-message --queue-url "https://sqs.eu-west-1.amazonaws.com/183742092277/platform-manager-reconciler-staging.fifo" --receipt-handle "AQEB3ftrEn12QA+2gjjsZNu+CTbSqUHIgpxcMg0O7uMWWas4uVEIIHKH/8LT/e9vKBGd+EkZ5Oe0rJRJ6FxSHNnRkWkAIcUuQG7+dyYyuz1OTysnMxiK3RsitLhaYhhju9KZBwC59JaYoPQkeZjxjMfD/H07W3JrPxlCXUCL7/7spS3Ud5v7Qf7BjtX/8yXKaWDkessEABVjYDSnHReBW44i6+2Z2ZAUN+3Vs0w2R4mwOzOfodPUSIBHPpA2NFsc92fQcQC8yBIEvn1rlZb4kSRIOHTGvqjnsdDdGGtz/WIILYvLfL5EQ45zd2GZ3jkLziwg"
    

    find queue id from NR log when picking up queue item

  • upload wheel to artifactory

    poetry publish -r bts-artifactory-crtxctl-local -u stephensont -p <API_KEY>
    

2023-08-03

  • 1-2-1
  • complete SBOM render
  • troubleshoot rdp cluster: now stuck on newrelic (also happening in alpha)

2023-08-02

  • poc for upgrade paths
  • CEIP-4175: identity facilitation around external dns
  • support Khush on Jenkins terraform update
  • build new rdp beta cluster

2023-08-01

  • record monthly metrics
  • consider automation of SLI using robot framework
    • irritating to install due to some things needing zscaler on (pip) others off (npm)
    • playwright is newer browser automation similar to selenium