What vendors changed

GitLabElevated 5xx Errors on GitLab.com

GitLab.com experienced elevated 5xx errors affecting CI/CD, Git Operations and Website, with major impact, from 2026-05-12 21:04 UTC to 2026-05-12 21:33 UTC (29 min). GitLab identified a frozen Gitaly node; the incident was resolved and error rates returned to normal.

WWorkOSEmails being delivered with empty copy

Some WorkOS AuthKit emails, including invitations, displayed empty copy in some email clients. WorkOS rated the incident major; it ran from 2026-05-12 19:14 UTC to 2026-05-12 20:29 UTC, lasting 1 h 15 min. WorkOS applied a fix and confirmed newly sent emails displayed properly in Gmail again.

WWorkOSElevated Webhook Latency

WorkOS reported major webhook delivery delays starting 2026-05-12 17:59 UTC, resolved at 2026-05-12 18:41 UTC after 41 min. The team identified the issue and implemented mitigations; webhook delivery was recovering by 18:21 UTC.

DDevin (Cognition)Devin Session Disruptions

DigitalOceanGradientAI: Agent Platform Playground Interaction Errors

DigitalOcean reported a major GradientAI Agent Platform Playground outage: all agent interactions could return “Something went wrong,” while API endpoints remained unaffected. The incident started at 2026-05-12 10:50 UTC and was resolved at 2026-05-12 11:37 UTC, lasting 47 min. A fix was applied and normal service restored; the cause was not disclosed.

UpstashFly.io Upstash Redis Service Distruption

A critical Fly.io Upstash Redis degradation caused intermittent hangs and elevated error rates for a subset of instances. It started at 2026-05-12 09:49 UTC and resolved at 2026-05-12 13:48 UTC, lasting 3 h 59 min. Upstash implemented a fix; its postmortem reported Redis processes stalling inside a logging syscall.

CClosePartial outage - connection issues and application errors

Close reported a major partial outage affecting some users loading the application, making and receiving calls, and accessing the API. It started at 2026-05-12 09:34 UTC and was resolved at 2026-05-12 11:17 UTC, lasting 1 h 42 min. A fix was implemented by 11:07 UTC; no cause was stated.

NNx CloudErrors and slow responses from the Nx API

HHexProjects not loading

Hex projects were not loading on the Main site in a major-impact outage starting 2026-05-11 21:23 UTC. The incident lasted 17 min and was resolved at 2026-05-11 21:41 UTC; Hex reported implementing a fix but did not state a cause.

UpstashFly.io Upstash Redis service distruption (FRA region)

Upstash reported a major degradation affecting some Fly.io Redis databases in FRA, with increased latency or timeouts. It started at 2026-05-11 15:05 UTC and resolved at 2026-05-11 18:22 UTC, lasting 3 h 16 min. The postmortem described Redis processes stalling inside a logging syscall; the supplied root-cause explanation is truncated.

Red Hat (incl. Quay.io)Quay.io Intermittent HTTP 502 on Pull/Push

NetlifyLatency in API & Builds

Netlify reported a major incident causing increased latency in its API and Build Pipeline from 2026-05-11 13:36 UTC to 2026-05-11 15:27 UTC, lasting 1 h 51 min. Fixes were implemented and monitored before resolution; customers could experience slower API responses and builds during this period.

GitLab500 errors accessing repositories

GitLab reported a major outage affecting Git Operations and Website, with 500 errors accessing repositories and Gitaly errors when pushing. It started at 2026-05-11 08:41 UTC and was resolved at 2026-05-11 11:16 UTC, lasting 2 h 35 min. GitLab restarted the Gitaly process to restore service.

Elastic CloudTemporarily unavailability of Serverless Projects Functionality

Elastic Cloud reported a major outage affecting multiple Serverless projects, including AWS - N.Virginia (us-east-1) Elasticsearch Availability. Search, Ingest and EIS may have been unavailable from 08:00 UTC to 09:40 UTC on May 11. The incident started at 2026-05-11 08:00 UTC and was resolved at 2026-05-11 11:11 UTC, lasting 3 h 11 min; Elastic stated the root cause was resolved.

CCrusoeManaged Inference Services Inaccessible - Foundry and BYOM

OOrbElevated errors across API endpoints

RenderCertificate issuance paused for Let's Encrypt

Render reported a major Custom Domains incident from 2026-05-08 20:22 UTC to 2026-05-08 21:06 UTC, lasting 44 min. Let's Encrypt paused all certificate issuance, primarily affecting wildcard domains and, rarely, non-wildcard custom domains using its certificates; issuance resumed and the incident was resolved.

DiscordIncreased API Errors

Discord reported a major API outage that prevented many users from starting sessions, beginning 2026-05-08 19:08 UTC and resolved at 2026-05-08 22:38 UTC, lasting 3 h 29 min. Users regained platform access during recovery, and Discord confirmed all critical functionalities had recovered for all users at resolution.

RetoolMotherduck Resource disabled for investigation

Retool disabled the Motherduck resource, affecting Resource Queries, in a major outage starting 2026-05-08 19:06 UTC and lasting 13 h 36 min. Retool identified and fixed a caching issue, restored the integration, and marked the incident resolved at 2026-05-09 08:43 UTC.

Elastic CloudIssues with Email Alerts

Elastic Cloud reported a major incident affecting email alerts for some customers, starting 2026-05-08 18:10 UTC and resolved at 2026-05-09 01:01 UTC, lasting 6 h 50 min. A solution was implemented and email delivery resumed by 05-08 23:11 UTC; the final update confirmed successful delivery with no monitored issues.

Sumo LogicProblem with Billing and Account Management in North America 1 (US1), North America 2 (US2), Europe 1 (EU), Europe 2 (DE), Australia (AU), Japan (JP), Canada (CA), Korea (KR), America Federal (FED), Zurich (CH)

Sumo Logic reported a major Billing and Account Management outage: Parent organisations could not manage Child organisations when their list included Mumbai (IN). It started at 2026-05-08 10:50 UTC and was resolved at 2026-05-08 11:39 UTC (49 min), with normal operation restored.

UpstashQStash US Region Service Disruption

QStash's US-EAST-1 service suffered a critical disruption starting 2026-05-08 09:46 UTC, lasting 22 min and resolving at 2026-05-08 10:08 UTC. Upstash implemented a fix; its partial postmortem links the incident to a scheduler bug introduced in US East (N. Virginia), triggered when a user with active schedules deleted their account.

SingleStoreOperational issue - Amazon Elastic Compute Cloud (N. Virginia)Event

SingleStore reported a major AWS Clusters outage in N. Virginia, with customers potentially experiencing increased error rates and latency due to an AWS issue. It started at 2026-05-08 09:39 UTC and was resolved at 2026-05-14 14:56 UTC, lasting 6 days 5 h; SingleStore cited AWS's resolution notification.

TTwingateRelays443 Down

RRudderStackEvent ingestion issue for Starter and Growth customers in USEvent

RudderStack reported a critical US event ingestion outage affecting Starter and Growth customers; Enterprise customers were not affected. It started at 2026-05-08 03:44 UTC and was resolved at 2026-05-08 04:02 UTC, with a computed duration of 17 min. RudderStack attributed the disruption to an AWS load balancer issue affecting its ingestion layer.

SingleStoreSinglestore Portal Down

SingleStore reported a critical portal outage affecting Cluster Management, starting 2026-05-08 03:21 UTC and resolved 2026-05-14 14:57 UTC, with a recorded duration of 6 days 11 h. The vendor reported mitigation at 05-08 05:23 UTC and monitored afterward; no cause or specific fix was stated.

RRetell AIAgent reading and publishing broken

RRudderStackEvent ingestion issue for Starter and Growth customers in USEvent

RudderStack reported a critical US event ingestion outage affecting Starter and Growth customers; Enterprise customers were not affected. It ran from 2026-05-08 02:47 UTC to 2026-05-08 03:02 UTC (15 min). RudderStack attributed the disruption to an AWS load balancer issue and reported resolution.

CockroachDBDegraded performance in one AWS us-east-1 availability zoneEvent

CockroachDB Advanced reported a major degradation linked to elevated AWS error rates and degraded performance in one us-east-1 availability zone; clusters remained available. The incident ran from 2026-05-08 02:13 UTC to 2026-05-08 21:53 UTC (19 h 40 min), when all clusters in the affected zone were restored to a fully operational state.

Sumo LogicProblem with CSE APIs, CSE UI, Data Indexing, Metrics Alerting, Metrics Collection and Tracing Collection in North America 1 (US1)Event

Sumo Logic reported a major degradation in North America 1 (US1) from 2026-05-08 01:27 UTC to 2026-05-08 04:05 UTC, lasting 2 h 37 min. A cloud provider availability-zone failure caused log, metrics and tracing delays, CSE UI slowness and API latency or errors; raw data ingestion continued. Normal operation was restored.

NeonDatabase connectivity issues with projects in US-East-1 Cell 6Event

Neon reported major Database Connectivity degradation affecting approximately 1,000 projects in US-East-1 Cell 6, with slow or stuck compute starts beginning approximately 00:20 May 8, 2026 UTC. The incident was recorded from 2026-05-08 01:24 UTC to 2026-05-08 03:58 UTC (2 h 34 min). Neon reported mitigation and resolution but stated no cause.

NNordLayerNordLayer CP and SMP, authentication flow outageEvent

NordLayer reported a major outage affecting authentication, Control Panel and Service Management Portal functionality, starting 2026-05-08 01:19 UTC and resolved at 2026-05-08 04:30 UTC, lasting 3 h 10 min. NordLayer linked the incident to an AWS outage; all affected resources were online by 04:18 UTC, with no further functionality errors observed at resolution.

CConfluentElevated error rates and latency on produce and consume operations on AWS us-east-1 AZ4Event

Confluent Cloud reported a major incident affecting produce/consume operations in AWS us-east-1 AZ4, with elevated errors, latency and lingering connectivity issues. It started 2026-05-08 01:14 UTC and resolved 2026-06-04 16:47 UTC, lasting 27 days 15 h. Confluent attributed the initial impact to EC2 power loss during a thermal event and confirmed full operation at resolution.

CCanvaSome users may have issues accessing CanvaEvent

Some users had intermittent issues accessing Canva in a major outage starting 2026-05-08 00:43 UTC and resolved at 2026-05-08 22:43 UTC, lasting 21 h 59 min. Canva attributed the disruption to an ongoing AWS service disruption, reported recovery, and monitored before marking it resolved.

SSalesforceHeroku Service Disruption: DataEvent

Heroku’s major Data disruption affected a subset of US customers with intermittent connectivity, elevated DB latency and degraded third-party add-on performance after an upstream provider’s partial outage. Recorded from 2026-05-08 00:25 UTC to 2026-05-08 17:17 UTC (16 h 52 min), it involved an availability-zone migration and fixes for remaining read-only databases; a later update confirmed full recovery.

ConvexIntermittent Impact from AWS OutageEvent

Convex reported a major outage affecting Business, Free & Starter, and Professional, with downtime and elevated errors caused by an AWS Availability Zone issue. It lasted 1 h 49 min, from 2026-05-08 00:16 UTC to 2026-05-08 02:05 UTC. Convex moved affected infrastructure out of the impacted zone and confirmed stable service.

DatadogDelayed Monitors NotificationsEvent

Datadog reported critical delays in monitor notifications and platform data beginning at 23:39 UTC (recorded start: 2026-05-08 00:06 UTC). It identified upstream provider issues; data processing and alerting were normal by 06:21 UTC. The incident was resolved at 2026-05-08 14:07 UTC, with a recorded duration of 14 h 0 min.

BuildkiteDelays in job dispatch, webhook processing, and outbound webhooks

Buildkite reported a major degradation delaying job dispatch, inbound and outbound webhooks, and build notifications. The incident started at 2026-05-07 22:45 UTC and resolved at 2026-05-08 00:13 UTC, lasting 1 h 28 min. All customer workloads recovered; the postmortem reported no data loss.

GreenhousePerformance Degradation

Greenhouse reported major performance degradation affecting Greenhouse Recruiting - Silo 1, with delays in job posting, scheduling, webhooks, imports/sync, reports and other async workflows. It started at 2026-05-07 22:23 UTC and was resolved at 2026-05-07 23:42 UTC, lasting 1 h 18 min. A fix was implemented at 22:51 UTC; no cause was stated.

Red Hat (incl. Quay.io)Quay.io HTTP 502 on Pull/Push

SSalesforceTableau Service Disruption: Explore, Home Page, Publishto Tableau Cloud

Salesforce reported a major Tableau Cloud disruption affecting Sign In, Homepage, Explorer, and Publish to Tableau Cloud on 1 instance. It started at 2026-05-07 15:50 UTC and resolved at 2026-05-07 16:25 UTC, lasting 35 min. The issue self-resolved without Salesforce Technology team action; the trigger remained under investigation.

MongoDB AtlasAWS Cluster DNS Changes Degraded Performance

MongoDB Atlas for Government had a major AWS cluster DNS degradation from 2026-05-07 15:17 UTC to 2026-05-07 22:15 UTC, lasting 6 h 57 min; cluster topology changes could be delayed or fail. MongoDB patched bad metadata causing ACME-related DNS issues and confirmed DNS and cluster updates were healthy.

DDevin (Cognition)Devin Session Disruptions

BuildkiteJobs not starting on hosted agents and agent-stack-k8s

Buildkite reported a major Agent API outage: jobs failed to start on hosted agents, agent-stack-k8s and other implementations using acquire job. The incident started at 2026-05-07 08:48 UTC and was resolved at 2026-05-07 09:37 UTC, lasting 49 minutes. Buildkite reverted a change that supplied stale environment variables to acquire job.

DDevin (Cognition)Devin Session Interruptions

CCanvaSome users might experience errors when trying to access designs

Canva reported a major incident affecting some users accessing designs, with a “Trying to reconnect” error warning that continued editing could result in lost work. It started at 2026-05-07 03:29 UTC and was resolved at 2026-05-07 04:37 UTC, lasting 1 h 8 min; Canva identified the issue but did not state a cause or specific fix.

XeroGlobal: Resolved – Platform stable, Post Incident Summary to follow

Xero reported a major global outage causing errors for customers using the platform, affecting accounting, payroll, tax and mobile services. The incident started at 2026-05-07 02:20 UTC and was marked resolved at 2026-05-15 07:52 UTC; Xero reported implementing a fix and monitoring stability, with normal platform use restored.

RenderHA Postgres Unavailability

Render reported a major HA Postgres outage affecting multiple regions and Workflows, starting 2026-05-07 02:00 UTC and resolved at 2026-05-07 02:54 UTC (53 min). Render believed only HA-enabled databases were affected; Workflows recovered by 02:38 UTC and all affected database instances were back online by 02:47 UTC, though some showed stale 'Unready' statuses.

OOmni AnalyticsPages returning `Application DB Error`

Omni Analytics reported a major outage in EastUsa and Ireland, with pages returning `Application DB Error` from 2026-05-07 01:44 UTC until resolution at 2026-05-07 03:11 UTC (1 h 26 min). A scheduled internal database table update caused full page errors; a fix completed the update safely.

CCanvaSome users might experience errors when trying to access designs

Canva reported a major incident affecting design access, with some users seeing “Trying to reconnect” and a warning that continued editing could lose work. It started at 2026-05-07 01:32 UTC and was resolved at 2026-05-07 02:11 UTC, lasting 38 min; a fix was implemented by 01:55 UTC.