> cat /var/log/failures.log
A comprehensive record of when things went wrong.
🟢 CLERK
Mar 11, 2026
🟠 Clerk's JWT Templates are throwing 422 errors whenever an organization ID shows up to the party. Engineers found the culprit hiding in plain sight—fix incoming. Session management is getting the short stick while we wait. (Auto-resolved: no updates in 24h)
🟢 COCKROACHDB
Mar 11, 2026
⚪ AWS me-central-1 had a very bad day when someone tripped over a power cable in the UAE data center, yeeting multiple availability zones offline. CockroachDB clusters in the region went full zombie mode while me-south-1 just limped along like it'd been hit by a truck. AWS is still picking up the pieces. (Auto-resolved: no updates in 24h)
🟢 GRAFANA
Mar 11, 2026
🟡 Grafana's alert rules decided to stop doing their job for some lucky prod-us-west-0 customers. Turns out monitoring your monitors requires the monitors to actually work. Engineers are now nervously refreshing dashboards looking for clues. (Auto-resolved: no updates in 24h)
🟢 NETLIFY
Mar 11, 2026
🔴 Netlify's infrastructure decided latency was a feature, not a bug. Builds are failing, sites are crawling, and the API is basically a very expensive sleep function. Engineers are 'investigating' which is tech speak for 'frantically Googling their own error logs'. (Auto-resolved: no updates in 24h)
🟢 NETLIFY
Mar 11, 2026
🟡 Netlify's infrastructure decided latency was the real feature—builds timing out, APIs sluggish, error rates doing backflips. Investigation phase initiated, which is corporate shorthand for 'we're refreshing dashboards and hoping.' Your deploys are now in a holding pattern. (Auto-resolved: no updates in 24h)
🟢 OPENAI
Mar 10, 2026
🟠 OpenAI's file download feature discovered that 'error' is just another way to say 'surprise'. Engineering team frantically Googling 'how to make bits stay in order' while users debate if corrupted files are AI's attempt at abstract art. (Auto-resolved: no updates in 24h)
🟢 WEBFLOW
Mar 10, 2026
⚪ Update: [RESOLVED] Reports of issues with form submission notification emails
🟢 WEBFLOW
Mar 10, 2026
🟠 Webflow's email notification system decided to go full silent treatment—form data arrives safe and sound, but the 'you got a submission!' alerts vanished into the void. Engineering team's playing detective while your inbox remains blissfully unaware of customer inquiries.
🟢 REPLICATE
Mar 10, 2026
🟡 Replicate's speediest model decided to test what 'schnell' means in reverse. Flux Schnell is limping along while engineers stare at dashboards pretending they know what went wrong. (Auto-resolved: no updates in 24h)
🟢 OPENAI
Mar 09, 2026
🔴 OpenAI's Deep Research feature decided LATAM users needed to think shallower. Engineering team in 'mitigation mode,' which is corporate for 'we're Googling the error message.' Partial outage means some regions get answers, others get 404s and existential dread. (Auto-resolved: no updates in 24h)
🟢 LINEAR
Mar 09, 2026
🟠 Linear's workspace creation feature decided to ghost users in the US. The team is now frantically reverse-engineering why their own product won't let people, you know, create workspaces. Turns out even issue trackers need therapy sometimes. (Auto-resolved: no updates in 24h)
🟢 PLAID
Mar 09, 2026
🟡 Plaid's financial plumbing seized up, but they're claiming it's 'mitigated' while nervously watching the monitors like hawks. Translation: the bleeding stopped, but they're checking for infections every 30 seconds. (Auto-resolved: no updates in 24h)
🟢 GRAFANA
Mar 09, 2026
🟠 Grafana's prod-us-central regions discovered they're allergic to accepting metric data. Two separate 10-15 minute episodes of elevated errors and latency. The irony: a monitoring tool that can't monitor itself. Engineers are now staring very intently at dashboards that won't accept new data. (Auto-resolved: no updates in 24h)
🟢 ELASTIC
Mar 09, 2026
🟠 Elastic's Docker registry and orchestration went fishing in us-central1, but they're back from vacation. Monitoring's still twitchy though, like a recovering addict at a coffee convention. (Auto-resolved: no updates in 24h)
🟢 SNYK
Mar 09, 2026
🟡 Snyk's vulnerability scanner couldn't find its way back online between noon and 12:25 PM GMT. Engineers claim the patient has fully recovered and they're definitely not just refreshing dashboards nervously. The irony of a security tool having security theater. (Auto-resolved: no updates in 24h)
🟢 ANTHROPIC
Mar 08, 2026
🟠 Claude Desktop got stuck in an infinite loop trying to find tasks that technically never existed thanks to DST shenanigans. The app decided 1:30 AM doesn't exist, so it looked for it forever. Update to 1.1.5749 and pretend this never happened. (Auto-resolved: no updates in 24h)
⚪ AWS
Mar 07, 2026
⚪ AWS S3 in EU-CENTRAL-2 spent nearly an hour literally unable to PUT or GET its act together. Turns out the bytes weren't having it. Someone unplugged the reassembly line, plugged it back in, problem solved. Apparently AWS stores data like IKEA furniture—occasionally needs angry intervention.
🟢 SUPABASE
Mar 07, 2026
🟡 Uploads and downloads in EU-Central-2 decided to move at continental drift speeds. Supabase is investigating why their storage suddenly developed commitment issues. (Auto-resolved: no updates in 24h)
🟢 GRAFANA
Mar 07, 2026
🔴 AWS S3 decided to take an unscheduled sabbatical, dragging Grafana's EU region down with it. The good news? S3 remembered how to work again around 20:03 UTC. The bad news? Your monitoring was blind while the arsonist was in the building. (Auto-resolved: no updates in 24h)
🟢 SUPABASE
Mar 07, 2026
🟡 Supabase locked down subscription modifications while their billing provider ghosted them mid-investigation. Users stuck on their current plans like it's a bad relationship they can't escape. (Auto-resolved: no updates in 24h)