Switch deployment pipeline from GitLab Agent kubectl to FluxCD image
automation. CI now only builds and pushes tagged images to Forgejo
registry; FluxCD detects new tags and updates the deployment manifest.
Fix Ecto.MultipleResultsError on /dashboard when a device has multiple
Preseem access point records by using limit(1) instead of Repo.get_by.
get_bulk, set, and get_multi used 'with' chains that skipped
close_socket when the operation failed, leaking file descriptors.
Over time this exhausted the fd limit causing EMFILE crashes.
Switched to case + try/after to guarantee socket cleanup.
- Activity feed: use schema modules instead of string table names in Ecto
queries to fix UUID encoding (Postgrex expected binary, got string)
- Changelog parser: remove plain hyphen from separator character class
to prevent splitting on date hyphens (2026-02-13)
- Fix broken route paths in dashboard (/discovery, /subscribers/trace, /config-changes)
- Fix insights empty state settings link (org-scoped route)
- Add underscores to all 86400 literals across 6 files
- Alphabetize aliases in search.ex and agent_channel.ex
- Extract changelog parser helper to reduce nesting
- Extract dashboard impact calculation to reduce nesting
- Refactor agent_channel config change detection (pattern match, extract function)
- Combine double Enum.reject into single call in trace.ex
- Fix line length in trace.ex search query
- Replace length/1 > 0 with != [] in trace_live
- Run mix format on all files
- ChangelogParser: parses priv/static/changelog.txt into structured entries
- ChangelogLive: timeline view with DaisyUI styling
- Route at /changelog in authenticated scope
- What's New link in org dropdown menu
- Group alerts by site (tower/location) with collapsible sections
- Show alert count badges per site, critical count highlighted
- Color-code alerts by severity and age (red >1h, orange >15m, yellow new)
- Add Resolve button with confirmation dialog
- Show subscriber impact from Gaiia inventory at site and alert level
- Filter tabs: Unresolved, Critical, All, Resolved with counts
- Sort options: Severity (default), Oldest First, Subscriber Impact
- Sites with most critical alerts and highest subscriber impact sort first
- Use DaisyUI collapse, badge, tab, and btn components
- Add summary stat cards (subscribers, MRR, device count, alerts)
- Add QoE summary with capacity bar (qoe_color, qoe_bg, capacity_bar_color)
- Device health grid with status dots and response times (status_dot_class, format_response_time)
- Active alerts panel with urgency icons (insight_urgency_class, insight_urgency_icon, time_ago)
- Config changes panel with size labels (change_size_label, change_size_color, time_ago)
- Preseem insights panel with urgency styling
- Keep existing latency chart
- Use DaisyUI card/badge components throughout
- Health status dot (green/yellow/red/gray) from worst-case check state
- Relative 'last seen' with color-coded staleness thresholds
- Response time badge from latest monitoring check
- Subscriber count badge from Gaiia inventory items
- Batch queries to avoid N+1: Monitoring.get_device_health_summary/1,
Monitoring.get_device_latest_response_times/1,
Gaiia.get_device_subscriber_counts/1
Replace inline cache with dedicated buildcache tag that stores all
intermediate builder layers. Unchanged stages like deps and compile
will be pulled from cache instead of rebuilt.
Add node:22-bookworm container image since ubuntu:latest lacks Node.js
required by JavaScript-based actions. Update checkout to v6 and
build-push-action to v6.
- Promote Insights to top-level nav at /insights with source/urgency/status filters
- Unified insights page shows all sources (Preseem, Gaiia, SNMP, System)
- Bulk select/dismiss support
- Integration pages now show sync schedule and next sync estimate
- Expose sync_interval_minutes as editable config field
- Add tab navigation between integration sub-pages
- Dashboard insights section links to /insights
- Integration 'Network Insights' link now points to /insights?source=preseem
- Add product roadmap (ROADMAP.md)
- Add get_impact_summary/1: aggregates current outage subscriber/MRR impact
- Add list_active_incidents/1: down devices enriched with Gaiia/Preseem data, sorted by MRR
- Add get_site_impact_summaries/1: per-site health with QoE, affected subs, MRR at risk
- Dashboard top stats: MRR at Risk and Subscribers Affected cards
- Active Incidents section: color-coded by severity, shows duration/subs/MRR/QoE
- Site Health Grid: green/yellow/red health dots based on uptime and QoE scores
Gaiia migrated all status fields from scalar enums to object types,
not just AccountStatus. Update BillingSubscription.status,
InventoryItem.status, and nested billingSubscriptions.status in the
accounts query to request { id name } subfields and extract the name
value in sync mappers.
Auto-generate HMAC-SHA256 webhook secret when enabling the Gaiia
integration. Display webhook URL, secret with copy buttons, regenerate
option, and setup instructions on the integrations page.
Gaiia API changed AccountStatus and AccountType from scalar enums to
objects with id/name/type subfields. Updated the accounts query to
request nested fields and the sync mapper to extract the name values.
