changelog
LogTide 0.9 release notes
8 releases in the 0.9 line, newest first. Three major features: uptime monitoring with public status pages (#152), log parsing and enrichment pipelines with grok and GeoIP, and fully customizable dashboards with 9 panel types (#151).
SSRF Fix in Alert/Sigma Webhook Delivery
Closes a sibling-gap left by the 0.9.6 SSRF hardening: the alert/Sigma webhook delivery path now routes through the centralized safeFetch guard instead of the bypassable inline filter (GHSA-7v53-pw6r-99vj, CWE-918).
- SSRF fix: alert/Sigma webhook delivery now uses the centralized safeFetch guard (GHSA-7v53-pw6r-99vj, CWE-918)
- DNS resolution + per-redirect-hop revalidation + full IPv4/IPv6 private/reserved range coverage on webhook delivery
- Read-back oracle closed: blocked targets are rejected before the response body is read
- MONITOR_ALLOW_PRIVATE_TARGETS still lets self-hosted deployments target internal endpoints
Security
- SSRF in alert/Sigma webhook delivery via the legacy delivery path (authenticated) (GHSA-7v53-pw6r-99vj, CWE-918): the 0.9.6 SSRF hardening added the centralized
utils/ssrf-guard.tsguard and wired it into the HTTP/TCP monitors and theWebhookProvider, but the actual alert/Sigma webhook delivery path was left on the old inline filter.sendWebhookNotificationinqueue/jobs/alert-notification.ts(reached by thealert-notificationsBullMQ worker for threshold, rate-of-change and Sigma-rule alerts) still ran a barefetch(webhook_url, …)guarded only by a string-basedisPrivateIP(). That check was bypassable: any non-dotted-quad hostname returnedfalse(no DNS resolution), so a domain whose A record points at127.0.0.1/169.254.169.254/ an internal host passed the filter and the resolved internal address was then connected; the barefetchused the defaultredirect: 'follow', so a public host that302s to an internal URL was followed straight there; and CGNAT (100.64.0.0/10), IPv6, IPv4-mapped IPv6,0.0.0.0/8and198.18.0.0/15were not covered. An authenticated org owner/admin who can create a webhook notification channel could therefore use the backend as a blind SSRF probe against internal services and cloud metadata, with a partial read-back oracle since a non-2xx internal response body was spliced into the alert-history error message. The guarded “Test” button already blocked the same URLs, confirming this as an incomplete-fix sibling-gap. Fix:sendWebhookNotificationnow routes delivery throughsafeFetch(url, init, { allowPrivate: config.MONITOR_ALLOW_PRIVATE_TARGETS }), exactly asWebhookProviderdoes (DNS resolution + per-redirect-hop revalidation + full IPv4/IPv6 private/reserved range coverage), mappingSsrfBlockedErrorto the existing “private/internal addresses are not allowed” error. The inlineisPrivateIP/BLOCKED_HOSTSfilter and the barefetchare removed, and blocked targets are now rejected before the response body is read (closing the read-back oracle). TheMONITOR_ALLOW_PRIVATE_TARGETSopt-in still lets self-hosted deployments target internal endpoints. Reported by tonghuaroot
SPA Frontend, SSRF & Cross-Tenant Hardening, OIDC Issuer Fix
Frontend now runs as a full SPA on adapter-node (kills hydration bugs), SSRF guard on monitors and webhooks, cross-tenant projectId validation on four routes, OIDC iss-parameter fix for Authelia, and an infinite-skeleton fix on search/traces/metrics.
- Frontend converted to a full SPA (ssr=false root cascade), removing 22 per-route ssr=false workarounds and a whole class of hydration bugs
- Security: SSRF guard on uptime monitors and webhooks, denied by default (MONITOR_ALLOW_PRIVATE_TARGETS to opt in)
- Security: cross-tenant projectId validation enforced on alerts/preview, alerts, monitors and sourcemaps routes
- OIDC login fixed for issuer-identifying providers (RFC 9207 iss param) like Authelia (#233, #234)
- Fixed infinite skeleton spinner on search/traces/metrics when no project had a data-availability flag
Changed
- Frontend now runs as a full SPA on
adapter-nodeinstead of doing SSR with client hydration. A singleexport const ssr = falsein the new root+layout.tscascades to every route, so the server only ships the empty app shell and the client renders from scratch. Eliminates the entire class of hydration mismatch bugs that had been accumulating across login, register, invite, public status page, the/auth/callbackpage and the various dashboard subpages, each previously patched with its own per-routessr = false. Server load functions and the Node runtime are untouched (the adapter, API proxying, env vars, BullMQ etc. keep working exactly as before); there are no+page.server.ts/+layout.server.tsfiles in the repo today so nothing had to change semantically on the server side. UX tradeoff: the public status page at/status/[orgSlug]/[projectSlug]now flashes the empty shell for one paint before the JS hydrates, which is acceptable for an authenticated-by-default product but should be revisited if SEO on the status page becomes a goal (a single-line overrideexport const ssr = truein that page’s+page.tsputs it back on SSR without affecting anything else) - Removed 22 redundant route-level
ssr = falsedeclarations (dashboard/+layout.ts,dashboard/admin/+layout.tsand 20+page.tsfiles across landing, login, register, onboarding, invite, status, auth callback and the dashboard search / metrics / alerts / monitoring / projects / sessions / settings subtrees) that had been added one by one as each page hit a hydration bug. The new root-level cascade makes them all dead config; deleting them removes the temptation to copy the pattern into new routes
Fixed
- Infinite skeleton spinner on
/dashboard/search,/dashboard/tracesand/dashboard/metricswhen no project in the org had its data-availability flag set, typically right after a user deleted the only projects that had been ingesting. The filter logic atsearch/+page.svelte:441-444(and the identical pattern attraces/+page.svelte:142-145andmetrics/+page.svelte:90-93) readconst logsProjectIds = availability?.logsand then branched onlogsProjectIds ? filter : fallback. WhengetProjectDataAvailabilitylegitimately returned{ logs: [] }the empty array took the truthy branch ([]is truthy in JS) and[].includes(p.id)excluded every project, so the displayed project list was empty,loadLogs()never fired, andhasLoadedOncestayedfalseso theSkeletonTablerendered forever. Fix: guard the truthy branch withlogsProjectIds && logsProjectIds.length > 0so an empty availability response falls back to “show all projects” exactly like the API-failure path (.catch(() => null)) already does. The 0.9.4 backend optimization that introduced the cached availability flags is unaffected; this is purely a frontend null-vs-empty conflation. Logs of already-hard-deleted projects on the TimescaleDB engine remain unrecoverable due toON DELETE CASCADEonlogs.project_id(tracked separately under the soft-delete projects epic) - OIDC login failed against issuer-identifying providers (e.g. Authelia) because the
isscallback parameter was dropped (#233, #234): the OIDC callback handler extracted onlycodeandstatefrom the provider redirect and rebuilt the callback URL from just those two, discarding everything else. Providers implementing RFC 9207 (OAuth 2.0 Authorization Server Issuer Identification) appendissto the redirect andopenid-client’sauthorizationCodeGrant()validates it, so the token exchange failed with “issuer parameter missing” and login never completed. The Fastify callback route now forwards the fullrequest.querythroughhandleOidcCallbackinto the provider, which replays every parameter onto the callback URL handed to the token exchange (soiss,session_state, etc. survive); the requiredcode/stateare always re-asserted from the validated values. Duplicated/array-valued query params are collapsed to a single value withsearchParams.setinstead of being appended, since OIDC authorization-response params are single-valued per RFC 6749 / RFC 9207, avoiding a malformed URL with duplicateiss/codereachingauthorizationCodeGrant. Covered by new tests across the route, service and provider layers, including the array-collapse and undefined-param branches
Security
- Cross-tenant project access via unvalidated
projectId(authenticated): several routes accepted both anorganizationIdand aprojectId, verified the caller was a member of the organization, but never verified that the supplied project actually belonged to that organization. Since project UUIDs are normal identifiers that appear in dashboard URLs and client API calls, a user could pair their ownorganizationId(membership check passes) with a victim’s knownprojectIdand reach another tenant’s data. Confirmed on four handlers:POST /api/v1/alerts/previewreturnedsampleLogs(time, service, level, message, trace ID) from the foreign project;POST /api/v1/alertsandPOST /api/v1/monitorslet a rule/monitor be scoped to a foreign project (the monitor case also surfaces on the victim’s public status page, which renders byproject_id); andGET/DELETE /api/v1/sourcemapslet a member list or delete another tenant’s source maps. Fix: a sharedprojectsService.projectBelongsToOrg(projectId, organizationId)helper (singleprojectslookup filtered on bothidandorganization_id) is now enforced right after the existing membership check on each of those routes, returning403when the project is foreign. The alert/monitor update paths don’t accept aprojectIdand the custom-dashboards panel pipeline already had an equivalentensureProjectInOrgguard at its choke point, so no change was needed there. Regression tests cover each handler - Server-side request forgery (SSRF) and internal port scanning via monitors and webhooks (authenticated): HTTP/TCP uptime monitors and webhook delivery executed user-supplied targets from the backend’s network with no meaningful destination validation. Monitor creation only checked that an HTTP target started with
http(s)://and that a TCP target contained:;checker.tsthen calledfetch(target, { redirect: 'follow' })andcreateConnection({ host, port }), so a registered user could point a monitor athttp://169.254.169.254/…,http://127.0.0.1,10.0.0.0/8, etc. and use the sanitized up/down result and timing to probe internal services. The webhook provider had only literal-string private-IP filtering (no DNS resolution, incomplete IPv6, followed redirects), leaving DNS- and redirect-based bypasses open. Fix: a centralized outbound guard (utils/ssrf-guard.ts) resolves hostnames and rejects loopback, private, link-local (incl.169.254.169.254cloud metadata), CGNAT (100.64.0.0/10), multicast and reserved IPv4/IPv6 ranges (including IPv4-mapped IPv6 and ULA/fc00::/7, link-localfe80::/10). TCP checks resolve-then-pin the socket to the validated address (closing DNS-rebinding between validation and connect); HTTP checks and webhook delivery follow redirects manually and revalidate every hop instead ofredirect: 'follow'. The guard runs both at monitor create/update time (immediate400feedback) and at execution time (authoritative, returns ablockedresult). Private/internal targets are denied by default; self-hosted deployments that legitimately monitor internal services can opt back in withMONITOR_ALLOW_PRIVATE_TARGETS=true, which also governs webhook delivery. Note: HTTPS does not yet pin the connected address against a custom dispatcher, so a narrow DNS-rebinding window remains for HTTP monitors (tracked for a follow-up); the reported direct-target and redirect-to-internal vectors are closed
Multi-Engine Metadata Filters, Error Notification Throttle & Dependency Security
Metadata filters now work on ClickHouse and MongoDB, error notifications are throttled per group to stop email storms, the notification-channels defaults endpoint accepts monitoring, plus 19 Dependabot advisories resolved.
- Metadata filters now translated on ClickHouse and MongoDB, not just TimescaleDB (#226, #224)
- Error notifications throttled per error group via a race-safe cooldown (ERROR_NOTIFICATION_COOLDOWN_MINUTES, default 15)
- notification-channels defaults endpoint now accepts the monitoring event type
- Migration 043 adds last_notified_at to error_groups
- Security: 19 Dependabot advisories resolved (8 high, 11 moderate)
Fixed
- Metadata filters were silently ignored on ClickHouse and MongoDB (#226, issue #224): the metadata filter operators (
equals,not_equals,in,not_in,exists,not_exists,contains) were only translated by the TimescaleDB query builder. The ClickHouse and MongoDB query translators never readparams.metadataFilters, so any log search or alert rule that relied on ametadata.*filter came back unfiltered on those engines (the filter appeared to do nothing). ClickHouse now translates each filter to a predicate over the JSONmetadatacolumn usingJSONExtractStringpaired withJSONHasto distinguish a missing key from an empty string (not_in/not_equalssplit on whether the key is present,containsusespositionCaseInsensitive). MongoDB builds one clause per filter keyed onmetadata.<key>, all wrapped in$andso repeated filters on the same key don’t overwrite each other, withinclude_missingcontrolling whethernot_equals/not_inalso match documents where the field is absent ($exists). Covered by new per-engine translator tests - Error notifications spammed one email per occurrence:
processErrorNotificationsent an in-app notification, email and webhook for every exception occurrence, suppressed only when the error group’s status wasignored. A high-frequency client error (e.g. a Svelteeffect_update_depth_exceededloop firing thousands of times) produced one exception row per occurrence and therefore thousands of identical alert emails. The job now throttles per error group: it atomically claims a notification slot viaUPDATE error_groups SET last_notified_at = now() WHERE status != 'ignored' AND (last_notified_at IS NULL OR last_notified_at <= cutoff) RETURNING id, so only the first occurrence inside the cooldown window notifies and the rest are skipped. The conditional UPDATE is race-safe (concurrent jobs serialize on the row lock and re-evaluate the predicate against the freshly written timestamp), so even a burst of thousands collapses to a single notification per window. Cooldown is configurable viaERROR_NOTIFICATION_COOLDOWN_MINUTES(default 15, set0to notify on every occurrence). Occurrence counts on the error group and in-app dashboards are unaffected monitoringrejected by the notification-channels defaults endpoint:PUT/GET /api/v1/notification-channels/defaults/:eventTypevalidated:eventTypeagainst a local Zod enum of['alert', 'sigma', 'incident', 'error']that was missingmonitoring, even though the sharedNotificationEventTypetype, the DBorganization_default_channelsconstraint (migration 037) and the service all support it. Setting a default monitoring channel returned400 Validation error(“received ‘monitoring’”). Addedmonitoringto the route enum so the five event types are consistent across the stack
Added
- Migration 043
error_notification_throttle: adds a nullablelast_notified_at TIMESTAMPTZcolumn toerror_groups, used as the per-group notification cooldown anchor (see the error-notification throttle fix above)
Security
- Resolved 19 Dependabot advisories (8 high, 11 moderate) by bumping direct dependencies and pinning patched versions through the root pnpm
overrides.protobufjs→7.6.1(kept on the 7.x line via>=7.5.8 <8; covers code injection, prototype-pollution gadget, unbounded-recursion DoS, unsafe option paths, crafted-field DoS, overlong UTF-8 in@protobufjs/utf8 >=1.1.1).kysely→0.28.17(bounded>=0.28.17 <0.29; JSON-path traversal injection inJSONPathBuilder.key()/.at()).svelte→5.55.9(>=5.55.7; SSR XSS via spread attributes and promise serialization, DOM-clobbering XSS,<svelte:element>ReDoS).@sveltejs/kit→2.61.1(>=2.60.1;query.batchcross-talk).fast-uri→3.1.2(path traversal + host confusion via percent-encoded segments).qs→6.15.2(DoS inqs.stringify),devalue→5.8.1(sparse-array DoS),brace-expansion→5.0.6(numeric-range DoS),ws→8.21.0(uninitialized memory disclosure).protobufjsandkyselywere deliberately held on their current major/minor lines (their “latest” is a breaking jump) while still landing on the patched release
Cached Data-Availability, Activity Overview Multi-Engine Fix & Page Titles
Data-availability endpoint now reads cached flags on projects (6s+ → <50ms on ClickHouse), activity_overview panel works across all engines, 14 dashboard pages get proper titles, plus uuid 14 and fast-xml-parser 5.7.3 security bumps.
- GET /api/v1/projects/data-availability rewritten to read cached has_X_at flags - 6s+ → <50ms on production ClickHouse
- Boot-time one-shot backfill for the new project flags, throttled and gated by system_settings
- activity_overview panel logs/log_errors series now work on ClickHouse and MongoDB via reservoir.aggregate
- 14 dashboard pages now set proper <title> instead of falling back to generic 'LogTide'
- Security: uuid 14.0.0 (CVE-2026-41907) and fast-xml-parser 5.7.3 (CVE-2026-41650) bumps
Changed
GET /api/v1/projects/data-availabilityrewritten to read from cached flags onprojectsinstead of fanning outN*3existence queries to the reservoir per call. In production (ClickHouse, ~70M rows) the old path scannedcount(*)andLIMIT 1probes across logs/traces/metrics for every project in the org, taking 6s+ on non-trivial orgs. The new path is a singleSELECT id, has_logs_at, has_traces_at, has_metrics_at FROM projects WHERE organization_id = $1plus an in-memory staleness filter againstorganizations.retention_days, returning in <50ms regardless of ingest volume. The response shape is unchanged ({ logs: string[], traces: string[], metrics: string[] })- Ingest paths mark data-availability flags fire-and-forget: after a successful batch in logs / traces / metrics ingest,
projectsService.markHasData(projectId, kind)updatesprojects.has_X_at. A module-scoped in-memory debounce (5 min per(projectId, kind)) prevents UPDATE spam on hot projects: at most one UPDATE per 5 min per pod per project per kind. Failures are logged and swallowed so ingest is never impacted; the boot-time backfill is the safety net
Added
- Migration 042
project_data_availability: adds nullablehas_logs_at,has_traces_at,has_metrics_atTIMESTAMPTZcolumns toprojects. No backfill in SQL (handled at runtime, see below) - One-shot backfill at server boot:
runDataAvailabilityBackfill()runs fire-and-forget afterapp.listen, guarded by adata_availability.backfilledrow insystem_settingsso subsequent boots are a no-op. It selects every project with all three flags still NULL and, for each, runs threeLIMIT 1probes (logs / traces / metrics) via the reservoir in parallel. Throttled in batches of 10 with a 50ms pause to avoid hammering ClickHouse/Timescale/Mongo at boot. On a deployment with 1000 projects this takes ~6s on ClickHouse or MongoDB, up to ~30s on TimescaleDB when most projects are empty (chunk scan confirming zero rows). Force a re-backfill by deleting the safety row - Multi-engine staleness rule: a project is reported as “has data” only if
has_X_at >= now() - organizations.retention_days. Handles the “data aged out by retention” case without needing a background worker. The same logic works identically on TimescaleDB, ClickHouse and MongoDB deployments because the flag lives in Postgres and the reservoir only backs the ingest and backfill paths
Fixed
activity_overviewdashboard panel showed emptylogs/log_errorsseries on ClickHouse and MongoDB: the fetcher inpanel-data-service.tsgated both the logs and the spans queries behindreservoir.getEngineType() === 'timescale', so on non-Timescale engines those branches were skipped entirely and the panel rendered with all four log/span counters flat at zero (onlydetectionsandalerts, which read from Postgres operational tables, kept working). The logs path now branches: TimescaleDB still uses thelogs_hourly_stats/logs_daily_statscontinuous aggregate with the rawlogsfallback, while ClickHouse and MongoDB go throughreservoir.aggregate({ interval: '1h' | '1d' })(same dual pattern already used bybaseline-calculator). Bucket boundaries fromtoStartOfHour/$dateTruncalign with the panel’s UTCbuildBucketTimes, so the keys collide cleanly.logssumstotalper bucket,log_errorssumsbyLevel.error + byLevel.critical. Spans / span_errors stay TimescaleDB-only becauseIReservoirhas noaggregateSpans()primitive, matching the existing convention intrace_latencyandtrace_volume- Missing
<title>on 14 dashboard pages: every tab opened on/dashboard,/dashboard/admin/{organizations,projects,users,settings}(list and detail),/dashboard/monitoring(list and detail) and/dashboard/projects/[id]/{alerts,overview,performance,settings}showed the generic fallback “LogTide” fromapp.htmlbecause no<svelte:head><title>...</title></svelte:head>was set on the page. They now follow the existingTitle - LogTideconvention used by the rest of the app, with dynamic titles where the entity is already in scope:{$activeDashboard.name}for the custom dashboards root,{org.name}/{project.name}/{user.name ?? user.email}/{monitor.name}for the four[id]pages (each with a sensible fallback while the loader resolves). The two pure-redirect pages (/dashboard/projects/[id]and/dashboard/settings, whichgototheir default sub-route on mount) intentionally keep no title since the destination sets one immediately
Security
- Bump
uuidto 14.0.0 (GHSA-w5hq-g745-h8pq, CVE-2026-41907, MODERATE):v3,v5andv6accepted a caller-providedbuf+offsetbut did not validate bounds, so a small buffer or an out-of-range offset produced silent partial writes instead of throwingRangeErrorlikev1/v4/v7do.uuidis pulled in transitively bybullmqandldapts(production) andbson/ioredis(dev only); both production consumers resolve to their ESM entry points from our"type": "module"backend, so the major bump from^11.1.0is safe (uuid 12+ dropped CommonJS, but we never load the CJS path). Added"uuid": ">=14.0.0"to the root pnpm overrides; resolves to14.0.0. Engines requirement of Node 20+ is already satisfied byengines.node: ">=20.0.0" - Bump
fast-xml-parserto 5.7.3 (GHSA-gh4j-gqv2-49f6, CVE-2026-41650, MODERATE):XMLBuilderdid not escape the-->sequence in comment bodies or the]]>sequence in CDATA sections, so user-controlled data flowing into either could break out and inject sibling XML nodes (XSS in browser-rendered XML, SOAP injection, RSS/SVG payloads, etc.). The fix lands in 5.7.0; we use it transitively only via@aws-sdk/xml-builderpulled in by@types/nodemailer(dev), but bumping the override from>=5.5.7to>=5.7.0clears the lockfile advisory. Resolves to5.7.3
Traces Live Tail, Dashboard Panels & Async Ingestion Buffer
Traces gain live tail (SSE), row expand, keyboard shortcuts and export; monitoring reorganized into tabs; user settings promoted to a full page; new trace_volume and activity_overview dashboard panels; opt-in ReservoirBuffered async ingestion with in-memory and Redis Streams transports.
- Traces live tail via SSE, inline span expand, j/k/enter/r keyboard shortcuts and JSON/CSV export
- Monitoring page split into Monitors / Incidents / Maintenance / Status page tabs with summary cards
- User settings promoted from dialog to full page at /dashboard/account
- New trace_volume and activity_overview dashboard panels using continuous aggregates with raw fallback
- Opt-in ReservoirBuffered async ingestion: in-memory (signal-wakeup) and Redis Streams transports with DLQ, circuit breaker and Prometheus metrics
Added
- Traces live tail (SSE): new
GET /api/v1/traces/streamServer-Sent Events endpoint polls the traces table once a second and emits newtrace_ids as they appear, filtered byprojectId,service(CSV) anderror. On the frontend the traces page now has a “Live tail” switch + rows-limit selector (50/100/200/500/1000, persisted inlocalStorage) in the filter bar; incoming traces are prepended and the list is capped at the chosen limit. ThetracesEventSourceis torn down ononDestroyand on every filter-key change - Traces row click-to-expand with inline span list: clicking a trace row now toggles an inline panel below it that fetches
GET /api/v1/traces/:traceId/spansonce (cached pertrace_idfor the session) and shows a compact table with Service / Operation / Kind / Duration / Status for each span. The “View” action still opens the full trace detail page - Traces keyboard shortcuts:
/focuses the Trace ID input,rrefreshes,j/kmove the selection down/up with scroll-into-view,enterexpands/collapses the selected trace. Registered viashortcutsStoreunder scopetracesand unregistered on destroy - Traces export (JSON / CSV): new Export popover in the filter bar produces a client-side download of the current list (disabled during live tail or when the list is empty). CSV covers start_time / service / operation / duration_ms / span_count / error / trace_id
- Monitoring entry in the command palette:
cmd+know listsMonitoringunder Navigation (icon +g oshortcut hint), matching the sidebar’s Detect group so the page is reachable without leaving the keyboard trace_volumedashboard panel: new panel type that plots span count over time (optional errors line), configurable time range (1h/6h/24h/7d) and optional service filter. Mirrors the shape of the existing logs “Log Volume” panel but for OTLP traces. Reads thespans_hourly_stats/spans_daily_statscontinuous aggregate first, falling back to the rawspanshypertable withdate_trunconly when the cagg is empty for the window (covers theend_offset=1hrefresh lag on freshly-ingested data). TimescaleDB-onlyactivity_overviewdashboard panel: unified multi-series timeline combining logs, log errors, spans, span errors, detection events and alert triggers on a common bucket grid (hourly for ≤24h, daily for 7d/30d). Each series is individually toggleable in the config form. Per-source queries run in parallel and prefer thelogs_*_stats,spans_*_stats,detection_events_*_statscontinuous aggregates for cost; if a cagg returns zero rows for the window the fetcher falls back to the raw hypertable (logs,spans,detection_events) for that source only, so the panel stays correct on freshly-ingested data without running a raw scan in the hot path. Alerts always come fromalert_history+alert_rules(no cagg exists for it)
Changed
- Monitoring page reorganized into tabs: what was a single 1000-line scroll with three unrelated sections (Monitors, Incidents, Maintenance) plus a status-page config block wedged at the top is now a
Tabs.Rootwith four tabs - Monitors, Incidents, Maintenance, Status page - each with its own header, its own primary CTA, and its own empty state. The Monitors tab gains a summary card grid (Total / Up / Down / Paused, each card clickable to filter the list by that status), a search input that matches name / target / type, and a “Clear filters” shortcut. Every row now has a one-click Pause/Resume button (previously buried in the edit form). Status-page visibility, password, slug and embed-badge config moved into the Status page tab. No changes to underlying API calls or permissions - Admin dashboard sidebar collapses on mobile: the 256px admin sidebar was previously always visible, taking up a huge chunk of the viewport on phones. It now hides below
lg:and opens as a fixed drawer triggered by a hamburger in a new mobile-only sticky header that also shows the current section name (Dashboard / Users / Organizations / etc). The drawer has a backdrop, is dismissed on nav-click or backdrop-click, and locks body scroll while open. Desktop behaviour unchanged - Dashboard footer wraps properly on mobile: was a single
justify-betweenrow that squeezed “LogTide / Alpha v0.9.2” against ”© 2026 LogTide · Documentation · GitHub” on narrow screens. Now stacks the brand block over the links block belowsm:(each block keeps its own horizontal flow withflex-wrap gap-x-4 gap-y-1), and “Documentation” shortens to “Docs” undersm:to fit on a single line. Container padding also drops frompx-6 py-4topx-4 py-3on mobile - User settings promoted from a dialog to a full page at
/dashboard/account. The cramped scrollableUserSettingsDialogmodal (profile + password + tutorial restart + danger zone stacked inside a 32rem-wide overlay) is now a proper route using amax-w-3xlcontainer with each concern split into its ownrounded-lg border bg-cardsection (Profile / Change password / Onboarding / Danger zone). The dropdown menu item still says “User Settings” but now navigates instead of opening an overlay.UserSettingsDialog.svelteis removed. The delete-account confirmation remains a nestedAlertDialogsince that one is genuinely destructive and modal-appropriate - Public status page polished for mobile and clarity: outer container shrinks from
px-4 py-10topx-3 py-6on phones (onlysm:and up gets the roomy desktop padding). The 45-day uptime bars drop theirmin-w-[6px]floor to4pxon mobile so the whole row fits comfortably on 320px viewports. The footer (“Last updated …” and “Powered by LogTide”) grows fromtext-[10px]totext-xswith a taller top margin, and “LogTide” is now styled astext-primary font-medium hover:underlineso the link affordance is obvious (previously only a hover recolor). Incident update timestamps/status labels also bumped fromtext-[10px]totext-xs. The monitor-type badge (HTTP/TCP/etc.) hides belowsmto free up the row for the name + uptime%. The password input switches from a fixedw-64tow-full max-w-xsso it never overflows narrow phones - Monitoring forms moved into modal dialogs: the previous inline create/edit monitor form, the “New incident” form, the “Post incident update” form and the “Schedule maintenance” form were expanding/collapsing sections that shifted the whole page layout when opened. They now open as centered
Dialogmodals withmax-h-[90vh] overflow-y-autoso tall forms (monitor create) scroll independently of the page. The per-row “Post update” Dialog is a single instance driven by ashowUpdateFormid + anupdatingIncidentderived lookup, replacing the one-per-row inline expand. Form state, validations and submit handlers are untouched - Traces filter UI reworked into the same two-row pill bar as the search page: row 1 keeps Time Range (
TimeRangePickerin a popover), the Live tail switch, the List/Map view toggle and the Export menu. Row 2 exposes every filter as its own always-visible pill - Project (single-select), Services (multi-select), Status (tri-state: all / errors / ok), Duration range (min/max ms inputs), Trace ID (direct-navigate to the trace detail page). Pills switch to thesecondaryvariant when a non-default value is set; “Clear all” appears when any filter is active. Mobile popovers usew-[Xpx] max-w-[90vw]so they never overflow the viewport.handleTimeRangeChangesyncs picker state back intotimeRangeType/customFromTime/customToTimeso the trigger label survives popover close/remount - Traces backend accepts multi-value
projectIdandservice(CSV on the query string) and two newminDurationMs/maxDurationMsbounds onGET /api/v1/traces. The cross-project access check loops over every project in the list.tracesService.listTraceswidens its input types tostring | string[]and forwards toreservoir.queryTraces, which already supported these filters - Search page paginator now uses the same ellipsis/windowed control as the traces page: shows up to four numbered buttons plus sentinel
…on each side whentotalPages > 7. Falls back to the previous “Page N” label when the backend doesn’t return a total (storage engines withouttotalin the query response). “Previous” / “Next” labels hide on<smscreens so only the chevron shows on phones - Empty states wait for the first load to complete before rendering on the search, traces, errors and security-incidents pages. Each page now tracks a
hasLoadedOnceflag flipped in thefinallybranch of its loader; until the first fetch resolves (success or failure) the page shows its skeleton instead of the “No Logs Yet / Start sending logs from your applications” onboarding card. Prevents the flash where an org that already has data showed the onboarding CTA during the initial render between mount and the first query. When a fetch legitimately returns zero rows, the page also distinguishes “nothing matches these filters” (with a Clear-filters shortcut) from “project has no data at all” (the full onboarding) - Search page filter UI reworked into a two-row pill bar: row 1 keeps Search + mode, a Time Range popover trigger, Live Tail and Export inline. Row 2 exposes every filter as its own always-visible pill (Projects, Services, Hostnames, Levels, Trace ID, Session ID, Metadata) - each pill displays the current value (e.g. “Service: api-gateway”, “Levels: 2”) and opens its own popover for editing. Pills switch to the
secondaryvariant when a non-default value is set so active filters are visible at a glance. A “Clear all” link appears at the end of the row when any filter is active. Popover widths usew-[min(Xpx,calc(100vw-1rem))]so they never overflow the viewport on phones. No semantic change to the underlying filter behavior (state, query params, live-tail flow, metadata apply/clear) - purely a layout/affordance pass replacing the previous six-column “Filters” Card - Auto-created Default dashboard now uses the new
activity_overviewpanel instead of the logs-onlytime_series“Log Volume” chart. Existing organizations keep their current dashboard untouched; the swap only affects orgs that don’t yet have a default (i.e. new signups or orgs where the default was deleted and re-seeded viaensureDefaultExists) - Opt-in async buffer for reservoir writes: new
ReservoirBuffereddecorator sits between the ingestion API and the storage engine. When enabled viaRESERVOIR_BUFFER_ENABLED=true, POST/api/v1/ingestreturns as soon as records are accepted into a shard-partitioned queue; a flush consumer pool drains to storage asynchronously. Two transports ship in the box: an optimized in-memory queue with signal-based wakeup (single-instance, not crash-safe) and a Redis Streams transport with consumer groups,XAUTOCLAIM-based stale reclaim, atomicMULTI/EXECnack, and a DLQ side stream (multi-instance, durable). Circuit breaker bypasses to sync ingestion when the buffer fills beyond a configurable pending threshold or after repeated flush failures in a rolling window. Off by default; seedocs/async-buffer/for when to enable and the per-engine benchmark table IReservoirshared interface: public type thatReservoirandReservoirBufferedboth implement, so downstream code (backend monitoring, etc.) can type againstIReservoirwithout caring which implementation is live- Prometheus-style buffer metrics:
reservoir_buffer_enqueued_total,reservoir_buffer_bypass_total,reservoir_buffer_flush_success_total,reservoir_buffer_flush_failure_total,reservoir_buffer_flush_duration_ms(histogram),reservoir_buffer_dlq_total,reservoir_buffer_breaker_state(0 closed / 1 open / 2 half-open). All labelled by record kind and shard where applicable - Warning at startup when buffer is enabled on non-Timescale engines: benchmarks show the buffer regresses p95 under saturation on ClickHouse and MongoDB (the bottleneck is the flush side, which the buffer cannot hide). The backend now logs a clear warning with a link to the docs when
RESERVOIR_BUFFER_ENABLED=trueis set together withSTORAGE_ENGINE=clickhouseorSTORAGE_ENGINE=mongodb - k6 load-test script
load-tests/buffered-vs-sync.js(pnpm --filter @logtide/backend load:buffered-vs-sync): constant-arrival-rate 100 req/s for 3 min, batch size 10, reports p50/p95/p99 + error rate. The file lives in the gitignoredload-tests/directory; only the npm script is checked in - Graceful shutdown now drains the reservoir buffer in parallel with Fastify
app.close(): previously serialized, so a slowapp.close()(BullMQ workers, websocket connections) could exhaust the container stop timeout beforeshutdownReservoir()ran. The new order calls them viaPromise.all, giving the buffer its fullRESERVOIR_BUFFER_GRACEFUL_SHUTDOWN_MSwindow to drain regardless of how longapp.close()takes - Redis client in the backend’s buffer path is now explicitly
quit()-ed on shutdown:RedisStreamTransportdocuments that it does not own the client, so the backend owns the lifecycle. Without this, the Redis connection was torn down by the process exit instead of closing cleanly - Buffer start-up failure is fail-fast: when
RESERVOIR_BUFFER_ENABLED=trueand the consumer pool cannot start (e.g. Redis down at boot), the backend logs a CRITICAL message andprocess.exit(1)instead of silently continuing with a non-flushing buffer RedisStreamTransporthardening:XPENDINGnow filters by consumer name inclaimStale, so delivery-count mapping is accurate even on busy streams with claims from other consumers;getStats.oldestPendingAgeMsnow tracks the oldest not-yet-delivered entry (viaXRANGEwith exclusive start past the PEL tail) to match the documented contract;nackusesMULTI/EXECso the DLQ write and the original ACK either both land or neither does;enqueueManyinspects pipelineexec()results and throws on the first error instead of swallowing them; default consumer name changed fromconsumer-${pid}to${hostname}-${pid}-${randomHex}to avoid collisions between containers with the same PIDFlushConsumerisolates partial failures by kind: the three ingest calls (ingest,ingestSpans,ingestMetrics) now run viaPromise.allSettled; the breaker records success only when all three succeed; metrics reflect per-kind reality; DLQ only counts the failed kinds on retry exhaustion. An optionalFlushLoggercan be injected; default falls back toconsoleFlushConsumerPoolrelaxes drain check topendingRecords == 0: previously also requiredinflightRecords == 0, which never converged when another backend instance held claims on the same Redis consumer group. Consumer-task crashes are now surfaced vialogger.errorinstead of being swallowed byPromise.allSettled. Drain poll bumped from 50 ms to 200 ms to reduce chatterInMemoryTransportuses signal-based wakeup:dequeueno longer polls every 50 ms; enqueues wake waiters directly via a per-shard resolver list, andstop()unblocks waiters promptly.enqueueManynow batches records into a singleArray.push(...entries)per shard plus one wake, eliminating the N-await bottleneck that made the in-memory buffer slower than sync on fast TimescaleDB workloads (p95 dropped from 4224 ms to 24 ms in the 3-min 100 req/s test)- Reservoir exposes
getEngine()on the concreteReservoirclass (not onIReservoir): replaces the(inner as unknown as { engine }).enginecast insideReservoirBuffered’s constructor. Kept out of the public interface so backend consumers typed againstIReservoircannot reach the engine and bypass the decorator ReservoirBuffered.ingest/ingestSpans/ingestMetricsreturn the correctIngestResultshape:{ ingested, failed, durationMs }instead of the previous{ inserted }masked byas unknown ascasts.durationMsmeasures the enqueue time, so the value reflects the real cost imposed on the caller
Metadata Filters, Admin User Creation & Org-Scoped Status URLs
Generic metadata filters across log search and alert rules, admin-driven user provisioning, editable org/project slugs, org-scoped public status page URLs (BREAKING), plus SIEM heatmap crash fix and three fastify/SvelteKit security bumps.
- Generic metadata filter builder with 7 operators, backed by GIN-indexed JSONB queries and shared with alert rules
- Configurable metadata.* columns in the log search table, persisted per project
- Admins can create users from the dashboard (#198) without enabling public signup
- BREAKING: public status page URLs moved to /status/:orgSlug/:projectSlug
- Security: fastify 5.8.5 and SvelteKit 2.57.1 bumps (3 CVEs, 2 HIGH / 1 MEDIUM)
Added
- Generic metadata filters in log search and alert rules: the search page and alert rule dialogs now expose a “Metadata filters” section backed by a new
MetadataFilterBuildercomponent. Supported operators:equals,not_equals,in,not_in,exists,not_exists,contains. Filters are applied server-side via a GIN-indexed JSONB query builder in reservoir; alert evaluation also runs the same matcher in-process so rules can fire only when a specific metadata field matches - Configurable metadata columns in log table view: users can add arbitrary
metadata.*keys as extra columns in the search results table via a “Columns” picker. The selected column set is persisted per project in localStorage so it survives page reloads - Create user from admin panel (#198): admins can now provision new accounts directly from
Admin → User Managementvia a “Create User” button, without having to temporarily re-enable public signup. Opens a dialog to set email, name, password and optional admin role. Backed by a newPOST /api/v1/admin/usersendpoint that bypasses theauth.signup_enabledgate and logs acreate_userentry to the audit log - Set a custom dashboard as default from the UI: the dashboard switcher now shows a clickable star next to each org-wide, non-personal dashboard; clicking it promotes that dashboard to be the org’s default. Backed by a new
POST /api/v1/custom-dashboards/:id/set-defaultendpoint that atomically unsets the previous default and sets the new one in a single transaction, respecting the existing partial unique index. Personal and project-scoped dashboards are rejected with a 400 - Editable project slug from monitoring page: the status-page settings card in
/dashboard/monitoringnow exposes a “Public URL slug” input that lets the user rename a project’s slug, with inline validation against a shared format check (lowercase alphanumeric + hyphens, 2-50 chars), reserved-word list (api,admin,dashboard,status,auth,login,signup,logout,_app,health), and per-org uniqueness. Conflicts surface as 409 with a friendly inline error; race conditions are caught at the DB layer via the new composite unique index - Editable organization slug from settings:
/dashboard/settings/generalslug field is no longer read-only; owners can rename the org slug with the same validation rules and global uniqueness, with a warning that any existing status-page links and embed badges will break
Changed
- Public status page URL is now scoped under the organization (BREAKING): page and badge URLs changed from
/status/:projectSlugto/status/:orgSlug/:projectSlug. Affects the public web page,/api/v1/status/:orgSlug/:projectSlug/badge.svg, and/api/v1/status/:orgSlug/:projectSlug/badge.json. No redirect from the old URLs. Anyone embedding the badge SVG/JSON or sharing a status-page link must update the URL to include the org slug. Migration040simultaneously moves project-slug uniqueness from global back to per-org, so two different organizations can now both have a project namedfrontendwithout auto-suffixing
Fixed
- Security dashboard crashed with “Cannot read properties of null (reading ‘toLowerCase’)” (#200): root cause was two parallel
unnest(mitre_techniques)/unnest(mitre_tactics)calls in the same SELECT list ofSiemDashboardService.getMitreHeatmap. PostgreSQL evaluates sibling set-returning functions in lockstep and NULL-pads the shorter array, so any detection event whose tactic and technique arrays had different lengths produced heatmap rows with a null tactic, which then crashedMitreHeatmap.abbreviateTacticon the frontend. The heatmap query now unnests techniques only and resolves each one to its canonical tactic via the sharedMITRE_TECHNIQUESmap (with sub-technique to parent fallback), eliminating the malformed pairs at the source. FrontendMitreHeatmapalso filters cells with null tactic/technique andDetectionEventsList.getLogLevelClassdefensively handles a nulllevelas belt-and-suspenders
Security
- Bump
fastifyto 5.8.5 (GHSA-247c-9743-5963, CVE-2026-33806, HIGH): body schema validation could be bypassed by prepending a single space to theContent-Typeheader. Parser and validator disagreed on how to trim the header, so the body was still parsed but the schema lookup returned no validator and validation was skipped entirely. Upgraded from^5.8.3to^5.8.5 - Bump
@sveltejs/kitto 2.57.1 (GHSA-2crg-3p73-43xp, CVE-2026-40073, HIGH):BODY_SIZE_LIMITcould be bypassed under certain conditions inadapter-node. Tightened the pnpm override from>=2.53.3to>=2.57.1 - Bump
@sveltejs/kitto 2.57.1 (GHSA-3f6h-2hrp-w5wx, CVE-2026-40074, MEDIUM): callingredirectinside thehandlehook with a location containing characters invalid for an HTTP header threw an unhandledTypeError, enabling DoS if the location included unsanitized user input
Security Hardening, Input Validation & Bug Fixes
SQL injection fix in querySpans, ReDoS protection for monitor assertions, 25+ bug fixes across auth, pipelines, SIEM, monitoring, and SSR safety, plus comprehensive input validation on all route params.
- SQL injection fix in querySpans sortBy/sortOrder for both TimescaleDB and ClickHouse
- ReDoS protection on HTTP monitor body assertion via safe-regex2
- UUID validation, positive-int clamping, and max-length guards across all route params
- PII salt race condition, SSE duplicate sends, and Sigma sync FK corruption fixed
- localStorage SSR crash guards and URL-encoding fixes in frontend navigation
Fixed
- Identifier pattern update/delete failed with “Organization ID is required” (#193): PUT and DELETE routes required
organizationIdas an explicit query param or API key context, but session-based auth never sets that field. Now falls back to the user’s first organization, consistent with GET and POST - Project rename failed with “Expected string, received null” (#195):
updateProjectSchemausedz.string().optional()which rejectsnull, but projects without a description sendnullfrom the DB. Changed toz.string().nullable().optional()and updatedUpdateProjectInputaccordingly - Pipeline create/preview/import returned generic “Validation error” (#193, #194): POST routes expected
organizationIdin the request body but the frontend sends it as a query param. Routes now merge the query param into the body before Zod validation. Frontend error messages now surface the first validation detail instead of the generic label - Invitation accept race condition: wrapping the membership check + insert in a transaction caused the
accepted_atupdate to roll back when throwing “already a member”. Split the early-exit path out of the transaction and added 23505 unique-constraint handling for true concurrent accepts - SSE live tail duplicate sends:
latestLogpicked the oldest entry from a DESC-sorted array instead of the newest, causing every poll to re-fetch and re-send all logs since the oldest timestamp. Replaced with a defensive max-time computation - Sigma sync corrupted alert_rule_id FK: fallback
alertRuleId || existing.idwrote the sigma rule’s own PK into thealert_rulesFK column when no alert was auto-created. Now omits the column entirely unless a new alert rule was just created - Exception log viewer returned empty results for org-wide error groups:
getLogsForErrorGrouppassed an empty string asprojectIdto reservoir when no project filter was set. Now groups log IDs by their exception’sproject_idand issues onegetByIdscall per project - ReDoS in HTTP monitor body assertion: user-supplied regex pattern was compiled with only a 256-char length limit, no catastrophic-backtracking check. Added
safe-regex2validation and a compile-error catch - OTLP int64 precision loss:
parseInt()silently truncated int64 attribute values exceedingNumber.MAX_SAFE_INTEGER. Unsafe values are now kept as strings in metadata - PII salt race condition fallthrough: if two workers raced on the first hash for an org, the loser could return an unpersisted local salt when the retry read failed, permanently desynchronizing PII hashes. Now scopes the catch to 23505 and throws on unexpected errors
- Monitor status fallthrough on missing row:
processCheckResultsilently skipped the entire state machine (no status update, no notifications) whenmonitor_statuswas undefined. Now re-reads from DB or creates a default row before proceeding - Test isolation: auth mode pollution across test files:
system_settingswas never reset between tests, so any test settingauth.mode='none'caused 12 unrelated “401 without auth” assertions to return 503. Added cleanup + cache invalidation to globalbeforeEach - Sigma detection tests used
sigma_idafter code migrated toid: fixture data still passedrule.sigma_idassigmaRuleId, which would not match the new.where('id', '=', ...)queries translateDeletedropped level filter in reservoir:pushFilterreturn value was discarded for the level condition, corrupting$Nparameter slots when both service and level filters were present- SQL injection in
querySpansvia sortBy/sortOrder: user-controlled strings were interpolated directly into raw SQL in both TimescaleDB and ClickHouse engines. Added explicit column/direction allowlists ingestSpanshardcoded::uuid[]for project_id: ignored theprojectIdTypeengine option, breaking text-based project IDs with a Postgres cast errorlocalStorageSSR crash in organization store: 7 directlocalStoragecalls without a browser guard would throwReferenceErrorduring server-side rendering. Addedbrowsercheck on all accesses- Invite token not URL-encoded in redirect:
goto(/login?redirect=/invite/${token})corrupted the redirect path for tokens containing+,=, or/. Now wraps inencodeURIComponent ruleIdnot URL-encoded in security dashboard navigation: inconsistent withserviceNameandtechniquewhich already usedencodeURIComponent- Missing UUID validation on monitoring route params: all
/:idhandlers passedrequest.params.iddirectly to DB queries without format validation. Added Zod.uuid()parsing on every route - Negative limit/days in monitoring routes:
Number("-1") || 50evaluates to-1(truthy), passing a negative value toLIMIT. Replaced with aparsePositiveIntguard that clamps to[1, max] - Missing UUID validation on status-incident route params: same issue as monitoring -
:idparams were used unvalidated in DB queries - SIEM comment body has no max length:
z.string().min(1)with no upper bound allowed arbitrarily large comment payloads. Added.max(10000) - Notifications and alerts limit/offset NaN passthrough:
parseInt("abc")returnedNaN, which was passed to Kysely.limit(NaN). Added safe integer parsing with fallback and max cap - Correlation referenceTime accepted invalid date strings: schema validated only
{type: 'string'}, sonew Date("garbage")flowed into Kysely WHERE clauses asInvalid Date. Addedformat: 'date-time'and a defensive 400 response - Log pipeline comment contradicted jsonb merge direction: code comment said “do NOT overwrite existing” but the in-progress fix had flipped the
jsonb ||operand order so pipeline fields now win. Updated comment to match actual behavior
Service Monitoring, Log Pipelines & Custom Dashboards
Three major features: uptime monitoring with public status pages (#152), log parsing and enrichment pipelines with grok and GeoIP, and fully customizable dashboards with 9 panel types (#151).
- Service health monitoring with HTTP/TCP/heartbeat checks and public Uptime Kuma-style status pages
- Log parsing pipelines with 5 built-in parsers, custom grok patterns, and GeoIP enrichment
- Custom dashboards with 9 panel types, drag-and-drop reorder, resize, and YAML import/export
- Scheduled maintenances and manual status incidents on public status pages
- Versioned dashboard schema with migration framework and cross-org isolation guards
Added
Service health monitoring and status pages (#152)
Proactive uptime monitoring with auto-generated public status pages.
- 3 monitor types: HTTP/HTTPS (configurable method, expected status, headers, body assertion), TCP ping, and heartbeat (alert when no ping received within grace window)
- HTTP config: per-monitor
httpConfigwith method, expectedStatus, custom headers, and body assertion (contains or regex) - stored as JSONB, validated via Zod - Per-monitor severity: incident severity is configurable per monitor (
critical,high,medium,low,informational) instead of hardcodedhigh - BullMQ-style polling: worker checks all due monitors every 30s, batched in groups of 20 concurrent checks via
Promise.allSettled - TimescaleDB storage:
monitor_resultshypertable with 7-day compression, 30-day retention, andmonitor_uptime_dailycontinuous aggregate refreshed hourly - State machine: consecutive failure tracking with configurable threshold, atomic incident dedup guard (
WHERE incident_id IS NULL), auto-resolve on recovery - Auto-incident creation: when failure threshold is crossed, a SIEM incident is created with
source: 'monitor'and linked viamonitor_id; notifications sent via existing email/webhook channels - Public status page (
/status/:projectSlug): Uptime Kuma-inspired design with 45-day heartbeat bar grid, per-monitor uptime badge, overall status banner, custom CSS tooltips, light/dark mode toggle - Status page access control: configurable visibility per project - disabled (default), public, password-protected, or org-members-only
- Scheduled maintenances: create maintenance windows with start/end times; active maintenances suppress monitor incident creation and display a banner on the status page
- Manual status incidents: create public incident communications (investigating → identified → monitoring → resolved) with update timeline, independent from SIEM incidents
- Heartbeat endpoint:
POST /api/v1/monitors/:id/heartbeataccepts both API key and session auth, rate-limited to 600/min - Project slugs: auto-generated from project name on creation, unique per org, backfilled for existing projects via migration
- Dashboard UI (
/dashboard/monitoring): monitor list with project selector, create/edit/delete forms with client-side validation, detail page with refresh button, uptime chart, recent checks, copy heartbeat URL - Monitoring navigation: added to sidebar under “Detect” section alongside Alerts and Security
Log parsing and enrichment pipelines
Define multi-step processing rules that automatically parse and enrich incoming log messages before they are stored.
- 5 built-in parsers: nginx (combined log format), apache (identical to nginx), syslog (RFC 3164 and RFC 5424), logfmt, and JSON message body
- Custom grok patterns:
%{PATTERN:field}and%{PATTERN:field:type}syntax with 22 built-in patterns (IPV4, WORD, NOTSPACE, NUMBER, POSINT, DATA, GREEDYDATA, QUOTEDSTRING, METHOD, URIPATH, HTTPDATE, etc.) and optional type coercion (:int,:float) - GeoIP enrichment: extract country, city, coordinates, timezone, and ISP data from any IP field using the embedded MaxMind GeoLite2 database
- Async processing via BullMQ: pipelines run as background jobs after ingestion - zero impact on ingestion latency
- Project-scoped vs org-wide: pipelines can target a specific project or apply to all projects in the organization; project-specific pipelines take priority over org-wide ones
- Pipeline preview: test any combination of steps against a sample log message and inspect per-step extracted fields and the final merged result before saving
- YAML import/export: import pipeline definitions from YAML with
name,description,enabled, andstepsfields; upserts (replace existing pipeline for the same scope) - In-memory cache:
getForProjectcaches the resolved pipeline per project for 5 minutes, automatically invalidated on create/update/delete - Settings UI (
/dashboard/settings/pipelines): list, enable/disable toggle, create, edit, and delete pipelines with live org-switch reactivity ($effectinstead ofonMount) - Step builder: interactive UI for adding, reordering, and configuring parser, grok, and geoip steps with per-type configuration forms
- Pipeline edit page redirects to the list when the active organization is switched, preventing stale-ID errors
Custom dashboards with configurable panels (#151)
User-built dashboards replace the previous fixed /dashboard page, with team-specific views across all observability domains.
- 9 panel types covering every data source: time series, single stat, top-N table, live log stream, alert status (logs/alerts), metric chart and metric stat (OTLP metrics with avg/sum/min/max/count/last/p50/p95/p99 aggregations), trace latency (p50/p95/p99 from spans), detection events (SIEM by severity), monitor status (uptime + response time)
- Panel registry architecture: adding a new panel type touches only 6 files (shared types, backend Zod schema, backend fetcher, frontend panel component, frontend config form, frontend registry entry); the renderer, container, store, and routes never need to change
- Drag-and-drop reorder via
svelte-dnd-actionwith optimistic local state and a single PUT save - Drag-to-resize with bottom-right pointer-event handle, snapping to grid units; constrained by per-type min width/height from the registry
- Responsive 12/6/1 column grid that collapses panels to 6 columns on tablet (640-1024px) and 1 column on mobile (<640px); stored widths are always in the canonical 12-col reference and scale proportionally
- Auto-created Default dashboard per organization, idempotent via Postgres unique-violation guard, replicating the previous fixed layout (4 stat cards + log volume + top services + top error messages) so existing users see no visual change
- Inline edit mode with toggle, no separate edit page; pending changes are kept in a snapshot and discarded on Cancel
- Per-panel configuration dialogs with type-specific forms (level toggles, intervals, aggregation pickers, percentile selectors)
- Dashboard switcher dropdown in the page header with personal/shared distinction, create, delete (default protected), import, export
- YAML import/export: dashboards round-trip through YAML for version-controlling alongside infrastructure code; import regenerates panel IDs and uses
JSON_SCHEMAto block JS-tag prototype pollution - Versioned JSON schema (
schema_version: 1) with a migration framework in@logtide/shared: each version writes aMigrationFnindexed by target version,migrateDashboardwalks the chain on every read; clamps out-of-range versions defensively - Cross-org isolation guard: every panel data fetch verifies that
config.projectIdbelongs to the requesting org, preventing data leaks via crafted YAML imports or stale references - Batch panel data endpoint (
POST /:id/panels/data): single round-trip fetches all panel data viaPromise.allSettled, individual panel errors do not fail the dashboard - Organization scoping: dashboards are org-scoped with optional
is_personalflag (only visible to creator) andcreated_bytracking; partial unique indexes prevent multiple defaults per (org, project) scope - Migration
039_custom_dashboards.sql: JSONBpanelscolumn with GIN index for future panel-type filtering, partial unique indexes for default scope guarantees
Fixed
- Status page slug collision:
getPublicStatusnow filters bystatus_page_publicflag instead of returning the first project matching the slug, preventing cross-org data leaks createMonitornot transactional: monitor andmonitor_statusinserts are now wrapped indb.transaction()to prevent orphaned monitorsmapMonitortyped: replacedanyparameter with properMonitorWithStatusRowinterface for compile-time safety- Org membership check optimized: monitoring routes now use a single
SELECT WHERE user_id AND organization_idquery instead of fetching all user orgs and scanning in JS - Redundant DB read eliminated:
processCheckResultnow receives status data from the already-fetched monitor object instead of issuing a second SELECT - Target validation on update: PUT endpoint now validates target format against monitor type (HTTP must start with
http:///https://, TCP must contain:) $derived.byfix: monitor detail page uptime calculation now uses$derived.by()instead of$derived(() => ...)for correct Svelte 5 reactivity@constplacement: replaced invalid{@const}inside<div>elements with{#if}/{:else}blocks for Svelte 5 compatibilityuptimePcttype coercion: PostgresROUND()returns numeric as string - status page now coerces to number before calling.toFixed()- Default
failureThresholdaligned: frontend form default changed from 3 to 2 to match backend default - Test setup cleanup: added
monitor_results,monitor_status,monitors,incident_alertsto globalbeforeEachcleanup