Skip to content

Releases: Anirudhx7/ollama-mesh

v0.14.4

Choose a tag to compare

@Anirudhx7 Anirudhx7 released this 06 Jul 10:01

Changelog

  • 893c2b0 Merge pull request #7 from Anirudhx7/fix/warmup-embedding-only-models
  • 66470d3 Merge pull request #8 from Anirudhx7/fix/schedule-node-stale-ref-and-json-errors
  • f28ba95 feat(install): add uninstall.sh, OS-agnostic service mode, and real health checks
  • 509a5d9 fix(admin,ui): stale schedule node reference + malformed JSON error bodies
  • 644211f fix(router): warm embedding-only models via /api/embed, not /api/generate

v0.14.3

Choose a tag to compare

@github-actions github-actions released this 06 Jul 05:42
1fe39b7

Changelog

  • eeb62cf Merge pull request #1 from Anirudhx7/seo-repositioning
  • c562e34 Merge pull request #2 from Anirudhx7/fix-demo-meta
  • de8c722 Merge pull request #3 from Anirudhx7/fix-shutdown-log-race
  • 0e9d9fe Merge pull request #4 from Anirudhx7/fix/lifecycle-and-installer-batch
  • cc2b12a Merge pull request #5 from Anirudhx7/feature/scheduler-timezone-clock
  • 1fe39b7 Merge pull request #6 from Anirudhx7/fix/demo-server-clock-stall
  • cda5550 chore: relicense Apache-2.0, reposition SEO, rewrite roadmap, add DCO
  • 80c5969 docs(changelog): reconstruct v0.11.0-v0.14.2 history + Unreleased
  • 7c4a86c feat(seo): set up IndexNow for instant Bing/Yandex re-crawl
  • 87ec356 feat: add admin timezone configuration and db persistence
  • 942a3e9 feat: make handleSystemInfo return clock in configured timezone
  • 4898267 feat: show UTC offset prefix on timezone options
  • 5376779 feat: show live ticking server clock in scheduler ui
  • 943964a feat: sort timezone select dropdown chronologically by UTC offset
  • acba165 feat: use standard Intl API for dynamic timezone listing
  • f891b8c fix(admin): Bad Gateway on Hugging Face model pulls
  • 7dc4a66 fix(admin): drain async request-log queue before store close on shutdown
  • 8c8f72f fix(demo): add meta description + noindex to the dashboard SPA
  • 1fb6712 fix(router): warmup only loading one of multiple models to VRAM
  • 2633abb fix(router,admin): pinned models were unloadable via manual/scheduled unload
  • cb575a1 fix(router,config,admin): same node registered twice under different names
  • a39c9f5 fix(scheduler): silently logged 'fired' when an action was a no-op
  • 2841bcb fix(ui,installer): live version display, self-detection, SERVICE=1
  • c278ddb fix: mock system-info on the static demo so the server clock stops loading forever
  • d4e48f7 seo: optimize landing page & README headers for Ollama Load Balancer, High Availability & Failover keywords
  • 4eaa8ed seo: restore premium developer title and hero headline to maintain original branding aesthetic
  • 704fe89 seo: showcase both Ollama and vLLM multi-runtime load balancing in title and main headings

v0.14.2

Choose a tag to compare

@github-actions github-actions released this 02 Jul 19:43

Changelog

  • 172040f Fix pages.yml: remove tag trigger to prevent duplicate-SHA deploy failures
  • 2c1f89f fix(gofmt): format internal/admin/admin.go to pass gofmt CI checks
  • cc3615d fix(website): hide scrollbars on code blocks and optimize step layout for mobile
  • 2fdf487 fix(website): replace ugly default scrollbars with custom thin scrollbars for pre.code and .table-scroll
  • f8978df fix(website): style code block copy buttons to prevent overlap and squishing
  • 913d5dc fix: resolve all 7 security, concurrency, performance, and UI issues identified in codebase audit

v0.14.1

Choose a tag to compare

@Anirudhx7 Anirudhx7 released this 02 Jul 17:41

Changelog

  • 16dc9c4 Deploy Pages via gh-pages branch
  • 47f591c Fix GitHub Pages deploy queue handling
  • e5c19ae Use git tag for Pages version
  • 2ec7cf8 pages fix

v0.14.0

Choose a tag to compare

@github-actions github-actions released this 02 Jul 15:13

Changelog

  • bac3d6c Step 1: persistent warm state with authoritative startup reconciliation
  • 3262abf Step 2A: bench subcommand - cold vs warm TTFT through mesh proxy
  • c25dccd Step 3: Router Decomposition - Pure refactoring: split router.go into placement.go, health.go, queue.go and create interfaces.go
  • 36c5f65 Step 4: Weighted Placement Scoring - Replace basic routing with multi-factor scoring, model pinning, and node cooldown
  • 7bbcca6 Step 5: Predictive Prewarming - implement in-memory transition ring buffer, prediction worker, time-of-day pattern prewarming, and accuracy metrics
  • 8b11fba Step 6: Correct Option 2 description in README.md to reference multi-runtime subnet probing and PROBE=1 flag
  • c63a784 Step 6: Expand install.sh discovery scan to detect vLLM, TGI, and llama.cpp runtimes in addition to Ollama
  • ade88eb Step 6: Move Quick Start section to the very top of README.md to optimize developer UX
  • efb3ab5 Step 6: Update install.sh and README to support installing without probing, separated behind START=1/--start flags
  • 7541568 Step 6: Update installer and Docker compose with Prometheus & Grafana telemetry provisioning
  • 6ea186f Step 6: Update pages.yml concurrency to cancel in progress deployments
  • 3a6cad9 Step 6: Update website landing page installation step with standard versus network-probe options
  • 3bac673 Step 6: Upgrade Go toolchain to 1.25.11 to patch standard library vulnerabilities in govulncheck
  • e8e07c0 Step 6: Upgrade install.sh to probe the entire local physical subnet for active Ollama nodes and register them in config.yaml
  • 7351a3b WIP: persistent warm state before startup reconciliation fix
  • 2faa290 feat(lifecycle): manual + scheduled model unload
  • b699922 fix(ci): gofmt formatting + enforce LF line endings via .gitattributes
  • cb54f84 fix(ui): Models header counted the whole catalog as warm models
  • fab1371 fix: bug-sweep fixes from live + static review (analytics persistence, others)

v0.13.1

Choose a tag to compare

@github-actions github-actions released this 01 Jul 20:04

Changelog

  • 5317949 chore(docs): update product dashboard screenshot in README
  • 9262ae8 chore: bump version to v0.13.1
  • 5da0a94 fix(admin): backfill hourly analytics from SQLite on startup
  • ea6db6c fix(admin): rate-limit admin login to prevent brute-force
  • edba917 fix(docs+website): improve light mode colors contrast and optimize mobile view padding
  • 18f540d fix(router): use container network IP for Docker auto-discovery

v0.13.0

Choose a tag to compare

@github-actions github-actions released this 01 Jul 18:10

Changelog

  • d93ba32 Revert "chore: ignore config-e2e-test.yaml"
  • 368d8d4 chore: bump version to v0.13.0
  • 61661ad chore: ignore config-e2e-test.yaml
  • ab29409 feat(lifecycle): B1 LRU eviction primitives + pinned models
  • e7e3c6a feat(lifecycle): B2 headroom-gated auto-eviction on the load path
  • fc7924c feat(ui): B3 pinned-models control in Warmup page
  • 60f4494 fix(auth): validate expires_at format in handleAddKey
  • 435530c fix(demo+docs): add fetchNodes demo branch, fix LIMITATIONS logo
  • 7245641 fix(fmt+ui): gofmt router files; sync ScheduleForm node state on load
  • a76cf4a fix(router): guard affinity deletions under write lock; add recover() to background goroutines
  • d77f330 fix(ts): add React import for JSX.Element type in Warmup.tsx
  • 89561b8 fix(ts): use React.ReactNode instead of JSX.Element
  • 8ec3aab fix(ui): paused section edit form clip bug, pencil button toggles edit
  • 7177143 fix(ui): schedule edit form styling, active/paused split, smooth animations
  • 1a88a1f fix(ui+backend): favicon serving, 4-node demo, schedule edit/pause UI
  • 32dac27 refactor(ui): collapse warmup + pinned model lists, unify pill style
  • af25889 refactor(ui): declutter Warmup page - merge cards, collapse schedule form
  • 727b6be refactor(ui): tab toggle for warmup vs schedules at page top
  • e209ae9 test(warmer): verify pingNode request body and drain keep_alive

v0.12.2

Choose a tag to compare

@github-actions github-actions released this 01 Jul 11:49

Changelog

  • 4d8bddd chore: bump version to v0.12.2
  • 78da45e feat(warmup): scheduling + per-node warmup dashboard (W2+W3)

v0.12.1

Choose a tag to compare

@github-actions github-actions released this 01 Jul 11:35

Changelog

  • f5b73e1 chore: bump version to v0.12.1
  • 10324b7 chore: pin Go toolchain to 1.24 in CI and Docker
  • 269152d feat(routing)+docs: gate session affinity; honest peer-monitor; Anthropic 501
  • 6c4c85d feat(warmup): real per-node warmup + keep-alive guard + residency metric
  • 9c98c2e fix(auth,router,admin): rate_limit=0 unlimited, warmup timeout, data races
  • b88680d fix(fmt): gofmt openai_endpoints_test.go and store.go
  • ddd4318 fix(routing+api): correctness hardening and OpenAI API compliance
  • 038117e fix(ui/demo): keep login within app base path; populate demo Users page
  • 5cb1d45 fix(website): pre-paint theme (no light-mode flash on mobile); fix version + license
  • 2d46f7a security(config,router,admin): block link-local/metadata node URLs (SSRF)
  • 11d1d83 security(config,store): auth on by default, stronger entropy, DB 0600

v0.12.0

Choose a tag to compare

@github-actions github-actions released this 30 Jun 20:02

Changelog

  • 12d9013 chore: bump version to v0.12.0
  • 3fe2c24 chore: go mod tidy after adding golang.org/x/crypto
  • 4e2be7c feat(auth): enterprise multi-user auth system
  • 143f4dc feat(auth): separate admin/user login paths, user portal, fix non-admin change-password
  • 38e3291 feat(users): soft-delete, reset password, live key/model selectors in approve modal
  • 1847b9f fix(fmt): gofmt internal/store/store.go
  • 17c7df4 fix(security): bcrypt passwords, DB-backed admin auth, SQLite schema hardening
  • f6096fe fixed indentations
  • 080e114 style: optimize landing page scheduler console for mobile viewports