Developer Advocate Role extension | @Simeon Warner | It is too early to already see the results of Patricks work. Poll: vote to continue through the fiscal year (at least 3 months) 3 month: 4.500 $ Extension for 1 year for a sustainable continuation of his work: $ 36,000 Extension through June 2025: agreed with 6 votes in person → will be send out to the absend members What are the goals we want to achieve with this engagement? Ask the dev teams, including EPAM (e.g. Kitfox) See Google Doc “Responsibility” for his tasks Community doesn't have an overview of non-EPAM development → old list is long outdated Can Patrick (with the help of the members) put together a recent overview of this? Action for Simeon: extend contract, ask TC about their thoughts, reach out to Mike as well Discussion about: what do we want to finance with the members money? Responsibilities will have clear measurements for success.
|
AWS costs | @Simeon Warner | Simeon’s thoughts on AWS: Historically, FOLIO has paid for AWS for some dev teams (including EBSCO/EPAM and TAMU) and EBSCO has provided significant technical support funding in addition to their membership that has offset a significant portion of this. FY25 projections starkly changed this situations both with planned reduced contributions and unexpected AWS cost increases: --- FY22: AWS: $107,777, EBSCO: $ 79,500 (74%) ($31,300 of this actually booked in FY23) --- FY23: AWS: $228,441, EBSCO: $109,000 (48%) --- FY24: AWS: $230,118, EBSCO: $109,000 (47%) --- FY25: AWS: $460,000, EBSCO: $ 30,000 (7%) (August 2024 projections!!!!) --- Per 2024-09-09 CC meeting, AWS spending temporarily capped at $30k/month ($360k/year) There was no forewarning of the huge increase in projected AWS costs from FY24 to FY25. The change is due to a a combination of ramp-up in the number of environments needed to support development for both Okapi and Eureka environments, and increased testing work. There was limited community process around the decision to go to Eureka. There was no awareness of or planning around cost changes associated with Eureka. Eureka is primarily driven by EBSCO's need to meet the requirements of the LC implementation project. There is broad community support for Eureka as a good long-term direction for FOLIO, and thus EBSCO's work on Eureka is of significant benefit to the community. There are, however, some community concerns about the timing of Eureka and how to manage/reduce the time that both Okapi and Eureka need to be supported. Long-term, I (Simeon) think that the cost of development environments specific to dev teams should be paid for by the same mechanism that pays for the team. This is already the case for development work done by Index Data, K-Int, German libraries. This is the only scalable path -- in the current model the FOLIO community is suddenly on the hook to pay for additional effort that may be unconnected with increased membership.
DIscussion What is EBSCOs thinking? Agreement on Simeon's proposal in general What about shared development if brought to the community? Should be possible as well Model would mean: AWS invoices would go to the “stakeholders” (here EBSCO) directly FOLIO has a limited available capacity for administration Are we creating a risk of forking? The of test enviroments might change with EUREKA in place and the new app formalization → just a small portion is needed in test env. scope LOC development will decrease in future: go live is more or less reached Do we need a transition plan? What do we want to message out? The cost increase is for the EUREKA test and UAT environment community would fund for shared infrastructure, like UAT but not individual development env. Okapi is still included in the community funding for this fiscal year
Moving forward, we propose to adopt these principles: the community funds truly shared infrastructure including CI/CD, UAT and end-to-end testing environments development teams fund their individual dev/test environments cost sharing for any significant changes in spending or direction can be joint funded, subject to advance discussion, mutual agreement, planning and budgeting see draft letter from CC
Vote on this proposal in person: 5x yes, 1x abstain -we will send this out to the absent members
|