Infa - Taking ownership

[bnaya-payoneer]

Formulate the process

• [ ] Time estimation
• [ ] Define Priorities
• [ ] Formalize acceptance of:
  • [ ] Have a Feature Acceptance committee
  • [ ] Internal features
  • [ ] Open Sources
  • [ ] Formalize collaboration with other teams (import their staff)
    • [ ] Infra Champions
    • [ ] Getting Aways
    • [ ] Handover process
• [ ] SRE/DevOps cooperation
  • [ ] Having In-House DevOps (like the domain dev-ops)

Define the roles

• [ ] Working together
• [ ] Define our North Star

Set Expectations

• [ ] Excellency ;-)

  • [ ] Define our standard (Test coverage, documentation, stress test, telemetry, static code analysis)
  • [ ] Definition of done (Discovery-POC/Dev/QA/Production)

• ✗ Kanban

  • Kanban with Azure DevOps

Main Features

• NuGet migration -> V2
• Cache
• Event Sourcing
• K8s
• Configuration Refresh Trigger (consider to rebuild configuration)
• Dynamic UI
• OTEL
• Static Code Analysis and code fixes

[bnaya-payoneer]

Infra 1:

• Eventbus

  • a. Improve stability )CPU issues, Redis timeouts)

  • b. NIFI replacement with Kafka connectors

  • c. Migrate Application events to Kafka – multi phase project (Migration plan should be with deadlines in the R&D)

  • d. Flink/flitering options – this will enable full support of application events migration, consumers improvement in general (will reduce the need in the eventbus rules + will benefit ISAPI)

• e. Create guidelines for using Kafka/Rabbit/EB

• f. AI Moti: Can we close Event bus for new events by default, and only open with approval – after verifying the actual needs and tech requirements.

• Secret manager

  • a. Akeyless VS Vault decision

  • b. Migrate from Vendor credentials service to the new solution (cloud and main platform solution!)

• FSS development

  • a. MUST decide if “Go Anywhere” or any other product with Ido Adam

  • b. Currently developing fairness feature

  • c. What is planned for phase 2 ?

  • d. Status:n18.7 – Phase1 - deployed to QA.

• Parameters enforcement

  • a. The goal is to block direct DB updates on a shared table using API.

  • b. Currently blocked by devops portal development

  • c. AI Moti: check what is the cycle of snapshots for the parameters table, to allow a quick rollback.

• Outbox Pattern

  • a. Waiting for a schema solution – at a POC stage for the solution.

• Future:

  • 1. Kafka VS Rabbit –

  • a. Need to check how we can increase consumers on Kafka topic above the number of partitions

  • b. Need to do a load test and see best practices for each platform

  • c. Need to establish guideline for usages

• Infra 2:

  • State Machine .Net 8
  • Elsa VS Mass Transient POC
  • Prepare guidelines for migration and recommend the solution to the entire R&D
  • Service catalog
  • Multi phases project with devops portal. MUST align schedule + finalize design and requirements
  • Configuration service
  • Hot reload feature – High Priority
  • Update history/ audit information accessibility
  • Resources
  • basic cloud support (does not support complex resources) – Need to deploy to PROD
  • Caching
  • Replace pubcomp – Create New NuGet for In memory + distributed cache over Microsoft interfaces. (stop supporting “clear all” feature)
  • Mark pubcomp as obsolete + consider marking it as blacklist in NLN or block in pipelines the usage in new services.
  • Make decisions regarding the migration of index cache (should not continue as a solution! This was supposed to be temporary)
  • Tech forum for the final solution
  • Event Sourcing (should be in infra 3 – depends on the timeline)
  • Not enough information at the moment, project started by Bnaya and Roma
  • features are waiting for the stress test on the cloud.
  • Split Storage provider (stream + views)

• Smart .NET 8+ Template

  • Has a clear definitions of features in the epic

• Log Standard – Finishing replacement

  • Stringify status ?

• Open API NuGet – handover from Bnaya + fill TC

  • Done ?

• Future:

  • 1. Mono repo – decide if still relevant with Architects (Roma + Orel)
  • Schema is defined, Bnaya will lead additions to the smart cli template.
  • 2. Data-Base isolation considerations

• Infra 3:

  • ISAPI

  • Has many planned features in the backlog

  • The priority is to improve stability, reduce the data loss cases

  • Second priority – insert new data sources

  • Conductor

  • Deploy conductor service latest version to PROD

  • Automation process

  • DR solution – MUST have a war room at the end to confirm the success of the solution

  • Archiving improvements

  • ReportAPI – Data API

  • Has clear first phase in the backlog

  • The second phase is to create queries without a data engineer - Need to develop with Vadim the design and requirements

  • ODS – Mongo POC

  • Waiting for decisions – currently handled by Nir Agai. Missing clear deadlines and goals.

  • Future:

  • Take ownership – DMPS (no clear deadline- depends on the team ability to take the project)

  • Take ownership – Files API (no clear deadline- depends on the team ability to take the project)

  • Event Sourcing – based on Capacity + need to find users before developing!

• General tasks for the group

  • Flink POC (can be in Infra 1 or 3 – based on availability) – The goal is to establish a plan for filtering ability (this is a blocker for the application event migration)
  • Need to verify that FLINK will be ready as an On Prem solution, and justify using Flink over Kstream/ KSQL options.
  • NLN – always upgrade
  • SNYK upgrades – all the services as well
  • .Net 8 upgrade – all services upgrade (validate after each upgrade that the NLN passes! – there are some gaps in Microsoft NuGets)
  • Improve alerts and dashboards – alone and with Omri Bardichev