2023-11-30
- RDP clusters
- escalaion to AWS for
rdp-tooling - extend platman timeout to 5 mins (but does not have effect due to other timeout somewhere) https://github.com/elsevier-centraltechnology/cortex-platform-manager/pull/576/files
- escalaion to AWS for
- MUST get to the self-assessment
- TIL / Learned: run reconciler from branched code.
2023-11-29
- GitHub actions
- talk Khush thru adopting
- fixed the issue with checkout of second repo
- potential enhancement discussion: https://global-elsevier.slack.com/archives/C010TPUE8JU/p1701262055267119
- RDP clusters x2
- SALSA workaround plus ongoing issues but eventually got past those
2023-11-28
- catch up post cluster steps
- docs:
- GHA for crtxctl
- ceip-4839:
- rewrite single PR, a mere 63 builds later… LOATHE
- adopt central runners
- ceip-4839:
- backlog:
- review http://localhost:58630/posts/logging-dashboard/.
- Can it be automated?
- animated gif to better demonstrate the steps?
- review http://localhost:58630/posts/logging-dashboard/.
- camunda 8
- helm charts: https://docs.camunda.io/docs/self-managed/platform-deployment/helm-kubernetes/overview/
- tasklist only included for non-production use: https://camunda.com/download/self-managed/kubernetes/
- can slack integration replace tasklist? maybe with grafana?
- install kind:
brew install kind - install camunda: https://docs.camunda.io/docs/self-managed/platform-deployment/helm-kubernetes/guides/local-kubernetes-cluster/
kind create cluster --name camunda8 kubectl cluster-info --context kind-camunda8 # write the sample nginx ingress yaml ## nginx_ingress_values.yml #controller: # replicaCount: 1 # hostNetwork: true # service: # type: NodePort helm install -f nginx_ingress_values.yaml nginx-ingress oci://ghcr.io/nginxinc/charts/nginx-ingress --version 0.18.0 wget https://github.com/camunda/camunda-platform-helm/blob/main/kind/camunda-platform-core-kind-values.yaml helm repo add camunda https://helm.camunda.io helm repo update helm install camunda8 camunda/camunda-platform -f camunda-platform-core-kind-values.yaml export RELEASE_NAME=camunda8 kubectl port-forward svc/$RELEASE_NAME-zeebe-gateway 26500:26500 & kubectl port-forward svc/$RELEASE_NAME-operate 8081:80 & kubectl port-forward svc/$RELEASE_NAME-tasklist 8082:80 & kubectl port-forward svc/$RELEASE_NAME-optimize 8083:80 & kubectl port-forward svc/$RELEASE_NAME-connectors 8088:8080 & kubectl port-forward svc/$RELEASE_NAME-web-modeler-webapp 8084:80 & kubectl port-forward svc/$RELEASE_NAME-web-modeler-websockets 8085:80 & kubectl port-forward svc/$RELEASE_NAME-keycloak 18080:80 &
- microk8s (canonical) vs k3s (suse)
2023-11-27
- RDP:
- EUC and GP on it
- Khush
- talk thru a change
- needs ability to write to PR -> me to work out demo in crtxctl
- also has grafana issues
- releases:
- ceip-4791-calico-ns-beta: https://github.com/elsevier-centraltechnology/cortex-platform-blueprints/pull/140
- reinvent:
2023-11-24
OKRs : minimal
RDP prod cluster
- confirm TPR2
- added to us-east-2 logging platform:
- confirm pre-reqs
- attempt; salsa failed
NewRelic test case
- basic smoke test done (for now it’s failing)
- start exploring NR api from python
2023-11-23
- OKRs, minimal thinking about
- PPE
- shutdown nodes (2 meetings)
- TPR3
- Khush
- grafana (no help)
- sonarqube (also patched some crtxctl issues)
2023-11-22: vacation
2023-11-21
- closed: cortex-operations PR review:
- jira review:
- closed https://elsevier.atlassian.net/browse/CEIP-4344: typha config map
- closed: https://elsevier.atlassian.net/browse/CEIP-4139: robot spike and sub-tasks
- tigera-calico namespace labelling
- pre-sync tested in ArgoCD dev
- CEIP-4789: prepare PRs for rollout: CEIP-4790 (copy this for the other two)
- cortex inspector role should mirror plat manager approach (prod manages nonprod)
- inspector namespace labelling
2023-11-20 2023-11-23
- OKRs
2023-11-22: vacation
2023-11-21
- closed: cortex-operations PR review:
- jira review:
- closed https://elsevier.atlassian.net/browse/CEIP-4344: typha config map
- closed: https://elsevier.atlassian.net/browse/CEIP-4139: robot spike and sub-tasks
- tigera-calico namespace labelling
- pre-sync tested in ArgoCD dev
- CEIP-4789: prepare PRs for rollout: CEIP-4790 (copy this for the other two)
- cortex inspector role should mirror plat manager approach (prod manages nonprod)
- inspector namespace labelling
2023-11-20
- tidy up:
- https://elsevier.atlassian.net/browse/CEIP-1903: small dependency doc
- https://elsevier.atlassian.net/browse/CEIP-3895: verify removal of old charts
- discussion of how to apply labels to calico before kyverno locks it out, with Ashish
2023-11-17
- PPE migration incident 12h!
2023-11-16
- PPE / Cloudwatch
- 33rd account with cross account log subscript breaks AWS 🤦
Capability testing PRs
- Advisor
- add second check into report and resolve bugs
- DKP
- Final word from Irfan? - do TPR
2023-11-15
- escalate PPE case to Felipe
- encourage CWS to roll instance types to prod
- encourage Khush about demo
- Capability Testing IAM
- Work with Ashish - 2h!
2023-11-14
- October metrics
- PPE AWS case
- advisor:
- docker build
- had to reinstall docker due to failed upgrade
- fixed advisor docker build but not reproducible for Khush, he fixed differently
- bunch of questions about KSI work from Felipe
- follow enablement threads as on duty, esp. cws (Thomas, Felipe dealing) and ipi (Ashish)
- finally back to local capability testing
- cannot any longer assume inspector agent role either from inspector, or from EnterpriseAdmin
2023-11-13
- vacation half day
- planning
- troubleshoot advisor, mentoring Khush
2023-11-10
- Incident and PIR: https://elsevier.atlassian.net/wiki/spaces/TIOCORTEX/pages/119601337991608/2023-11-10+Cortex+Postmortem+cws-cws-prod
- Advisor: shorter conversation with Khush
- Karl: prep slides for achitecture case sudy
- BTS: showcase Argo
- new tickets:
2023-11-09
Tech Showcase
From AI to Eternity by Professor Wendy Hall
- didn’t work?
NPS, permit repeat submissions
build ppe-cpt-prod cluster
2023-11-08
- Tech Showcase
- Okta
- more than just social login (reset pwd etc)
- lots of tailoring via existing SDKs for all sorts of platforms
- passkeys
- rather than second factor like yubikey there is no password, starts right from account creation
- support from applpe, google, ms thru fido org
- Okta
- Mute ppe-cpt-prod till
- Khush 2-3h?
- Dev10: Flowfest 2-6pm
2023-11-07
- Tech Showcase
- Reaxsys
- Need continuous consistency so considered Cassandra but had to rule out!!
- ~2b rows in Oracle -> ETL to spark then HBase
- 25k events = ~15h (5x faster)
- data equally distributed by using hash keys
- Ethical Hacker: Glenn Wilkinson
- Use canary tokens for early warning
- phish.io??
- Reaxsys
2023-11-06
- Tech Showcase
- Guy Kingham: Elsevier Life Sciences
- falling return on investment coming together with meaninful GenAI
- biology knowledgebase to identify morbidity characteristics followed by chemistry database to identify solution
- MEH
- Sam Whiting: Agile is not autonomous
- Agile manifesto does not support exept with self-organising teams, which is challengable
- innovation thrives in constrained environments, changing one thing following scientific method
- autonomy privileges path of least resistance and therefore becomes de-motivating
- should not be told to expect or even want autonomy, instead seek space to learn (experiment)
- Guy Kingham: Elsevier Life Sciences
- PPE
- prod cluster request
- raised
2023-11-03
- CEIP-4469: KSI migration
- impact analysis
- tickets
- do it
- more capability tests?
- fluentbit interesting to Matteo
- ADR for capability testing IAM
- catch up with Khush
2023-11-02
- 1-2-1
- 2024: the Heroku-like offer
- how to get embedded in partner teams to gather real world requirements
- need to get working on okr eval
- standup and enginneering forum
- demo capability testing
- chat with Khush
- X typo in report_manager_helper_functions.py
- X no tests! (wants to, need to setup project)
- X git repo
- X poetry
2023-11-01
- 1.25 readiness checks.
- unfortunately it seems we were not interested in resolving IPI’s concern but just running the previous script and manipulation for prod.
- a good half day wasted, learn not to dive in too quickly when ‘just’ need xyz