Sprint 143 Planning: Fabric8 Analytics

[sivaavkd]

OSIO Analytics Sprint 143

Sprint Goals

Publishing IDE extensions for OSIO analytics platform (VSCode, DevStudio)

• UX related enhancements for the stack report (Stack Report design revamp #1773)

• Identify performance bottlenecks and publish matrix for OSIO resource requirements based on load testing results (Scale OSIO analytics platform  #1144)

• Deploy 3scale APICast gateway, Update IDE workflow to include 3scale user registration and test IDE workflow end-to-end with 3Scale gateway deployed - to PROD. (Deploy self hosted 3scale gateway  #1451, VSCode integration with 3scale registration portal #1285)

• List and close all the marketplace related TODOs (icons, READMEs) for publishing IDE extensions ( Items related to publishing VSCode extension to marketplace #1455 )

Analytics recommendations for Vert.x and SpringBoot and quick-starters

• Recommendations for Vert.x and SpringBoot (Improve companion recommendations for entire Vert.x and SpringBoot ecosystem to support release of IDE extenstions #1727)
• Spikes to try out supporting Node ecosystem ([Spike][PoC] Deep and Wide models to support Node ecosystem #1736 and Generate stack recommendations using a probabilistic approach with Edward #1802 )

Testing / Quality related

• UI Integration tests UI integration tests for Analytics UI #1693
• enable/improve E2E tests for ingestion components: [13] Perform integration tests for data ingestion related components of OSIO analytics #1329

Getting started experience

• Based on feedback, develop 2nd leg for POC for Analytics driven flow(POC for Dependency Editor flow #1764)

CVEs

• ([3] Document steps how to map package identifiers to CPEs (used in NVD data) #1264)

• Periodically get up-to-date CVE data for Maven components from NVD #1767

Improve license recommendation for end user

• Determine a representative license for a github repository given by user #1761
• Add all the licenses that are unknown to license analysis platform #1695

Bugs

• Incorrect usage outliers show up at stack report #1429
• Bug: Show the unknown license count for unique unknown licenses in the stack report UI #1147
• PGM scoring fails with OOM error #1490
• fabric8-analytics API server uses half a CPU core even though there is no traffic [prod-preview] #1463

Feedback Loop

• Add user input stack to PGM training. Account for User Input stacks as part of recommendation training #1694
• Save user's feedback for the stack analysis report Capture user feedback for the stack-analysis report. #1740

Stretch goals (POCs, pulled in from H train priorities)

• Display Insights from Fabric8 analytics at Che IDEs UI as POC (POC - Display Insights from Fabric8-Analytics at Che UI #1768)
• POC for CVE alerts on Pipeline view (POC for CVE alerts on Pipeline view #1770)

Engineering work

Work on high priority Technical Debt

• Expose workers bookkeeping data: [8] Expose workers bookkeeping data via API #1404
• Able to figure out and sync the delta between OLAP and OLTP stores: Able to figure out and sync the delta between OLAP and OLTP stores #1085
• Remove logging from PGM once it is tested on production: Remove logging from PGM once it is tested on production #1785

[krishnapaparaju]

@sunilk747 Please include the License related EPIC (next steps) as part of the sprint plan

[krishnapaparaju]

@miteshvp Please include PGM caching related work item

[krishnapaparaju]

@invincibleJai Please include work items related to quick POCs around (a) Insights at Che (b) Flagging CVEs, Stack level Licenses at build, pipelines screens. Goal is to iterate a quick POC and get it reviewed with PM team.

[krishnapaparaju]

@rootAvish @abs51295 Please include spikes around research around analytics: (a) based on Bayesian networks with Edward to have Bayesian networks run on deep neural networks (b) Plain deep neural networks

[krishnapaparaju]

@samuzzal-choudhury we wanna wrap up fixing performance bottlenecks by end of sprint # 143. Please change #1144 / create a new issue and include as part of the sprint plan.

[krishnapaparaju]

@miteshvp Please include the issue covering remaining work items related to insights for VertX and SpringBoot ecosystems.

[krishnapaparaju]

@sivaavkd please remove #1521

[krishnapaparaju]

Please remove code merge related work until UX work is done for this activity

[krishnapaparaju]

@sivaavkd From this sprint, please remove integrating into Che files as currently OSIO support Che version 5x and this Che files format related work is being planned for Che version 6x.

[bmicklea]

This is a very aggressive sprint - perhaps we should start listing priorities in the description here? From my perspective the most important things are:
#1773
#1144
#1264
#1767
#1761
#1695
#1693
#1329

[krishnapaparaju]

@bmicklea Ack. Doing well with these work items marked as priority. I have included an new icon to indicate priority and maybe would be good to review during spring demos..