HackerTrans
TopNewTrendsCommentsPastAskShowJobs

takinosh

no profile record

Submissions

[untitled]

1 points·by takinosh·27 giorni fa·0 comments

[untitled]

1 points·by takinosh·3 mesi fa·0 comments

[untitled]

1 points·by takinosh·3 mesi fa·0 comments

[untitled]

1 points·by takinosh·4 mesi fa·0 comments

Okd-metal-installer – Bare-metal Kubernetes via static ISOs

github.com
2 points·by takinosh·4 mesi fa·2 comments

Show HN: A Kubernetes Operator to Validate Jupyter Notebooks in MLOps

github.com
2 points·by takinosh·5 mesi fa·0 comments

MCP ADR Analysis Server

tosin2013.github.io
1 points·by takinosh·11 mesi fa·0 comments

Universal AI Development Standard – A Common Protocol for AI Dev Tools

github.com
2 points·by takinosh·anno scorso·1 comments

comments

takinosh
·4 mesi fa·discuss
I built okd-metal-installer to make bare-metal OKD installs less dependent on fragile PXE-style boot infrastructure.

The tool uses Ansible inventory plus group_vars / host_vars as the source of truth for cluster topology and host configuration. It generates the required Ignition config, compiles static network configuration using NMState, and customizes a CoreOS ISO per host with those settings embedded.

Instead of relying on network boot, the boot media can be delivered either manually or through Redfish virtual media. The deployment flow in the repo is:

preflight validation

Ignition generation

network config generation

ISO customization

config serving

DNS configuration

boot delivery

Current support includes:

single-node, compact, and HA topologies

disconnected / air-gapped installs

Route 53 or internal dnsmasq for DNS

FCOS / SCOS support

Repo: https://github.com/tosin2013/okd-metal-installer
takinosh
·anno scorso·discuss
UAID-001 is a research project that defines a universal protocol and directory structure that AI development tools can adopt to provide consistent developer experiences, enable seamless tool-switching, and encourage shared context across tools.

This project aims to solve the fragmentation problem in the AI-assisted development ecosystem by creating a standard that multiple tools can implement, allowing developers to work across tools without losing context or functionality.