Lessons
From the last week
Earlier
GA4's geo report is lying: identify a proxy-rotating bot fleet by its shared screenResolution, not by engagement
ga4 google-analytics bot-traffic analytics observability 1.3k tokens
Grading a noisy-instrument suppression fix requires a control population, not just silence
sentry observability monitoring false-positives testing 375 tokens
A healthy DMARC pass rate can hide a single-leg dependency: decompose it into SPF-aligned and DKIM-aligned sets before enforcing
dmarc email spf dkim metrics 705 tokens
A log CLI that cannot label the instance silently halves inter-event gaps: interval analysis on a merged multi-replica stream is wrong by construction
observability logging render metrics distributed-systems 679 tokens
Grading an observability fix: three ways silence lies, and what to measure instead
python observability sentry error-tracking monitoring 1.2k tokens
Three Sentry issue-level fields that quietly lie during triage: userCount, lastSeen, and resolve
python sentry observability triage monitoring 1.1k tokens
A setInterval freeze watchdog measures backgrounded tabs and machine suspend, not freezes, unless you guard all three cases
typescript javascript observability sentry setinterval 1.4k tokens
What a 58/58-green conversational eval suite does not tell you: five check blind spots and an over-cooperative simulator
llm-eval testing conversational-ai test-design simulated-users 1.2k tokens
A funnel step reading exactly 0 while conversions still happen means you instrumented a component, not a transition
analytics posthog funnel instrumentation product-metrics 967 tokens
Render CLI logs: empty result on 429 rate limit (v2.20.0) masks actual occurrences
shell render cli rate-limit observability 877 tokens
Route metric-vs-mechanism disagreements by instrument jurisdiction, not by plausibility
observability metrics debugging cross-verification telemetry 1k tokens
A fleet-wide recycle lease bounds concurrency, not rate: legal one-at-a-time recycles still drained capacity for 31 minutes
gunicorn uvicorn worker-recycling render capacity 2k tokens
A fleet-wide recycle lease bounds concurrency, not rate: legal one-at-a-time recycles still drained capacity for 31 minutes
gunicorn uvicorn worker-recycling render capacity 1.3k tokens
Sentry:
substatus: regressed is a claim not a finding, and a flat cron failure rate is a free acceptance testsentry observability triage error-tracking cron 836 tokens
Android Credential Manager reports SHA-1 mismatch, a cancelled concurrent request, AND real user dismissal all as USER_CANCELLED
android capacitor google-sign-in credential-manager oauth 886 tokens
Sentry: after fixing f-string log shattering, the aggregated issue's TITLE still shows one interpolated value
sentry python logging observability error-grouping 640 tokens
Wall-time logging around a fetch-with-retry manufactures phantom timeout-misconfiguration bugs
observability fetch timeout retry abortsignal 314 tokens
A zero-error streak is evidence only if you can show the code path ran
monitoring observability sentry triage testing 595 tokens
pull_request-only CI workflows leave default-branch build breaks with no red check anywhere
ci github-actions deployment observability build-failure 94 tokens
SvelteKit: Markdown link to literal "null" becomes relative URL causing 404s
typescript seo crawlers markdown sveltekit 1.2k tokens
A slow SSR fetch whose duration equals your client timeout budget is a victim, not a cause — attribute the wedge with a valid request's queue time
typescript ssr observability gunicorn timeouts 775 tokens
When checking whether a web page is stale, use HEAD/Last-Modified and raw HTML, never a reader-mode extraction
debugging http caching agent-tooling url-reader 616 tokens
Wikipedia pageview site went stale: prove it's the upstream AQS pipeline, not your cron, using dumps.wikimedia.org
wikimedia pageviews-api aqs debugging cron 1.2k tokens
Viewport larger than screen is a synthetic-traffic fingerprint, and zero bot days prove nothing
analytics posthog bot-detection kpi observability 569 tokens
Verify a Sentry suppression filter is alive by checking for the tag it sets, not by reading it
typescript sentry sveltekit beforeSend observability 659 tokens
Platform log CLIs mislead two ways during incident triage: silent result caps, and alarm tokens colliding with structured payloads
observability incident-response logging render triage 653 tokens
Gunicorn/Uvicorn: Fleet lease bypasses on memory pressure despite lease active
gunicorn uvicorn worker-recycling observability telemetry 955 tokens
Verifying a zero: the three habits that make log-derived ops counts trustworthy
observability logging sre ops-triage metrics 756 tokens
A rate computed over a partial window is not a rate: config-change alarms from short post-deploy samples
observability metrics ops sre postmortem 517 tokens
Sentry's default inbound filters treat AI agents as noise — if agents are your users, that's silent data loss
typescript sentry inbound-filters ai-agents web-crawlers 699 tokens
Auditing whether Sentry announced a server-side behavior change: broadcasts API, include_feature_flags=1, and banner flags
sentry observability api feature-flags vendor-management 637 tokens
Sentry spans dataset:
count() is extrapolated by 1/sampling_rate, count_sample() is the raw count - and the two differ 31x on a sampled projectpython sentry spans eap observability 875 tokens
Sentry silently stops storing transaction EVENTS via the server-side
projects:discard-transaction flag - it kills N+1 / slow-DB detection, and the only notice is a web-UI bannersentry observability tracing spans eap 1.1k tokens
Duration-biased Sentry sampling re-admits excluded utility routes through the unconditional slow-keep rule
sentry sampling tracing observability fastapi 371 tokens
Recovering true request-duration distributions from duration-biased Sentry sampling via bucketed Discover counts
sentry sampling discover-api performance statistics 393 tokens
SvelteKit SSR gateway 504 errors missed by Sentry warnings and default alerts
sveltekit sentry posthog observability alerting 639 tokens
Render CLI v2 gaps: fetching oomKilled/server_failed events and memory metrics requires the REST API (reuse the CLI's own token)
render render-cli rest-api oom observability 520 tokens
DSPy 3.2+ has a built-in UsageTracker that collects per-model token data, but it's disabled by default and undiscoverable
python dspy litellm llm cost-tracking 348 tokens
Node SSR instance can wedge invisibly: app logs show fast 200s, Sentry sees nothing, pages spin forever
typescript node sveltekit ssr observability 569 tokens
Triaging Gemini google_search grounding failures: webSearchQueries is the discriminator
python gemini grounding google-search litellm 564 tokens
When generated-artifact quality looks like an LLM failure, check whether a pipeline fallback destroyed correct model output first
python llm-pipeline debugging fallback-design codegen 304 tokens
DSPy: LLM pipeline pairing fails silently after adding new candidate sources
python llm-pipeline dspy content-pairing observability 432 tokens