Live outage feed

Outages and recoveries from the services you depend on.

Earlier incidents

40 incidents

Back to latest

Sep 16, 2026

  1. AkamaiPerformance issues in HTTPS edge delivery network

    We can confirm that the issue was mitigated at 00:57 UTC on 15 September 2026 and the service has resumed normal operation. Customers and partners can view additional details about the incident by logging in to: https://community.akamai.com/customers/s/feed/0D5a700001TpkinCAB or reaching out to Akamai Support. We apologize for the impact and thank you for your patience and continued support. We are committed to making continuous improvements to make our systems better and prevent a recurrence of this issue.

    Resolved

  2. TelnyxOutbound Messaging Latency

    The elevated latency affecting messaging outbound requests has been resolved. We are continuing to monitor the service.

    Resolved

  3. SumUpMFA Outage on Latam Bank Flows

    The issue has been fixed and we’re confident it’s resolved SumUp services have returned to normal. You can stay up to date on the status of our services anytime here on our status page: https://status.sumup.com.

    Resolved

  4. New RelicData Irregularities for EU Region

    Between 23:06 UTC and 23:16 UTC, some customers in the EU region may have experienced errors when sending metric timeslice data to New Relic, issues in viewing metric timeslice data in real time, and delayed or missing alert notifications related to real time data. Impacted data may not be recoverable.

    Resolved

Sep 15, 2026

  1. GitHubDegradation in repos contents API

    On September 4, 2026, between approximately 21:45 and 22:07 UTC, some users experienced errors and elevated latency for repository operations. The incident was fully resolved at 22:23 UTC.<br /><br />The cause was a capacity change that spread one of our clusters across additional availability zones; our zone-aware traffic routing kept sending requests to the original zone for performance, overloading a small set of servers while the new capacity sat idle. We resolved the incident by reverting the change and letting traffic rebalance.<br /><br />We are improving per-zone capacity guarantees, cross-zone load-shedding, and pre-production testing of multi-zone changes to prevent recurrence.

    Resolved

  2. GitHubIncident with several GitHub Services

    On September 13, 2026, between 08:43 and 10:44 UTC, GitHub experienced degraded availability across approximately 28 services, including Issues, Pull Requests, Actions, Codespaces, Pages, Notifications, Code Scanning, Git LFS, and new account signup. At peak, 8.8% of requests to create GitHub App installation access tokens failed. Token issuance for Actions workflows was also affected, impacting approximately 4% of workflows during the incident time frame. Creating issues through the web interface failed for about 96% of attempts, and signup failures were above 90%. <br /> <br />The cause was an internal data-cleanup job that began writing to a shared database cluster at 07:33 UTC. That cluster stores permission data read on nearly every authenticated request. The safeguard that was pacing the background job watched only one health signal — how far the database replicas were lagging — and that signal stayed low the whole time. It did not account for the load building on the primary itself, so the job kept writing while the primary quietly ran toward its limit. <br /><br />When the primary ran out of available connections, requests that needed it could not complete. First, there was no quick timeout on these database calls, so request handlers waited on the stalled database instead of failing fast, and the shared request-handling capacity degraded into site-wide errors. Second, a retry loop around token creation kept re-sending the writes that were already failing, which held the database saturated rather than letting it recover. <br /><br />Monitoring declared the incident at 08:50 UTC, but due to the broad impact and amplification from token creation, it took time to identify the source of the load. First responders mitigated by shedding internal load and pausing the job, and all services recovered by 10:44 UTC. <br /><br />To prevent recurrence, we are rate-limiting background jobs against shared, customer-serving databases by default, and adding automatic pausing and paging on primary-server load rather than replication lag alone. We are also surfacing running background work directly alongside database health signals so responders can see and pause it without leaving those dashboards, bounding retries in the token-issuing path, and adding request-level timeouts so one unhealthy database cannot consume shared web server capacity. In addition, we are breaking apart this database cluster to remove the single point of failure. We will be moving various service-specific data, including the authorization data, out of this shared cluster in the next two weeks.

    Resolved

  3. Box[Minor] Issue with Admin Reporting

    _Notice and disclaimer: Box is providing this preliminary information subject to further review and analysis. To the best of our knowledge, this is the current state and we may update as more information is confirmed._ On September 3, 2026 between 3:30 PM PDT and 7:12 PM PDT, some customers may have experienced missing, delayed, or inaccurate results within admin reporting features, including AI usage reports, in Box. Reports may have shown empty fields or data that did not reflect recent activity. The issue was caused by an outage in a third‑party data service that interrupted our data ingestion pipelines. Service availability was restored after the provider mitigated the problem. We are conducting a full engineering postmortem and our overview is subject to change with further analysis and findings. In the event our continuing investigation yields further substantive findings, we may publish additional analysis after the conclusion of the full engineering postmortem. We are continuously working to improve Box and want to make sure we are delivering the best product and user experience we can. We hope we have provided some clarity here and we would be happy to answer any questions you may still have regarding this matter. Sincerely, The Box Team

    Resolved

  4. DigitalOceanCloud Control Panel and API

    ## 1. What happened Between 16:56 UTC and 23:09 UTC on August 24, 2026, customers using DigitalOcean’s Cloud Control Panel, Public API, and dependent product management systems experienced elevated request error rates, peaking at 25%, in two distinct windows totaling 3.5 hours of active degradation within a 6-hour span. Running Droplets, network routing, and block storage \(Volumes\) remained fully functional and unaffected throughout the event. No customer data loss occurred. The event was initiated during a planned infrastructure migration of the database layer supporting identity, authentication, and authorization services. The migration introduced increased execution latency for a critical authorization query. In response to the elevated query latency, multiple upstream services initiated automated hedged retries. This secondary request amplification led to queue buildup in excess of the authentication service pool's processing throughput, resulting in cascading request timeouts across the authorization path. To mitigate the initial degradation, engineering teams rerouted authorization traffic onto a single healthy processing path and temporarily disabled hedging for primary callers, successfully restoring service availability. However, upon returning traffic to the standard multipath routing configuration, lingering query latency combined with retry behavior from unhedged callers caused recurrent queue saturation, leading to the second window of degradation. ## 2. Impact * Active customer accounts \(<2%\) experienced one or more API or Control Panel request failures during the event windows. * Aggregate failure rates across both impact windows averaged approximately 10% of authenticated requests, with a peak error rate of 25% during the second window. * Affected management control planes included the Cloud Control Panel, Public API, Kubernetes cluster authentication, Container Registry, and Serverless / AI Inference endpoints. * Data plane components—including active Droplets, network routing, and block storage volumes—maintained 100% availability, as established connections do not depend on the affected authorization control path. There was zero data loss. ## 3. Current state All impacted control plane services have been restored to nominal operational capacity. The database infrastructure migration has been completed. Remediation controls have been validated to ensure system stability against similar query latency scenarios. ## 4. Preventing recurrence To prevent future occurrences and enhance systemic resilience, DigitalOcean has implemented the following technical enhancements: 1. Automated Request Throttling and Circuit Breaking: Deployed automated circuit breakers across upstream services to detect downstream dependency degradation and automatically suppress hedged/retry amplification. Completed a broad architecture audit to ensure retry suppression and circuit breaking are uniformly enforced across all control plane services. 1. Queue Management and Load Shedding: Implemented fast-fail load-shedding mechanisms to prevent request queue saturation during dependency slowness, accompanied by proactive telemetry alerts to identify queue backing prior to customer impact. 1. Graceful Service Degradation: Configured critical authorization components to operate in degraded execution modes utilizing asynchronous, eventually-consistent data replicas during primary database latency events. ## 5. Timeline | **Time UTC** | **Event** | | --- | --- | | 16:56 | A specific authorization query began running significantly slower than its historical baseline. | | 16:58 | Automated monitoring detected abnormal error rates and latency; detected within 5 minutes of impact starting. | | 17:05–18:52 | Operators tuned retry configuration and consolidated requests to a healthy service instance \(first unavailability window closed\). | | 21:30 | Shifting traffic back to the production pattern caused the same saturation problem \(second unavailability window begins\). | | 21:30–23:09 | Teams disabled the retry/amplification behavior across the remaining affected services and added additional read capacity for the slow query. Service was restored to a stable state. |

    Resolved

Older incidents