Doordash 장애 기록

1. 2026년 7월 24일 PM 5:15 UTCresolved

   약 04:00와 04:15 사이에 PT 7 월 24, 2026, 업스트림 인프라 제공 업체 (Cloudflare) 정전 드라이브 주문 제작 및 webhook 배달에 영향을 미치는 서비스 중단 발생. 이 ~15 분 창 도중, 몇몇 순서 및 상태 webhooks는 실패되거나 연기될지도 모릅니다. 업스트림 공급자가 복원 된 한 번 완전히 복구 된 서비스, 수동 완화는 DoorDash Drive 측에 요구되지 않았습니다. 우리는 붕괴를 위해 사과합니다.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 6월 28일 PM 10:33 UTCinvestigating

   We are currently investigating this issue.

2. 2026년 6월 28일 PM 10:45 UTCinvestigating

   We are continuing to investigate this issue.

3. 2026년 6월 29일 AM 12:26 UTCmonitoring

   A fix has been implemented and we are monitoring the results.

4. 2026년 6월 29일 AM 12:27 UTCmonitoring

   We are continuing to monitor for any further issues.

5. 2026년 6월 29일 AM 1:15 UTCresolved

   This incident has been resolved.

6. 2026년 7월 2일 AM 12:20 UTCpostmortem

   On June 28, 2026, DoorDash experienced an internal infrastructure issue that disrupted order transmission to POS-integrated merchants. As a result, Marketplace orders could not be delivered to affected merchant systems for approximately 90 minutes \(3:10 PM–4:40 PM PT\). The issue was identified, mitigated, and service was restored. During recovery, orders resumed processing successfully with operational safeguards in place, and normal service levels were fully restored by approximately 7:00 PM PT. We are implementing automated lifecycle management, completing a review of remaining legacy-managed infrastructure, and strengthening monitoring governance to improve early detection and reduce the likelihood of similar incidents recurring.

1. 2026년 6월 16일 PM 2:14 UTCinvestigating

   We are aware of the issue and are currently investigating

2. 2026년 6월 16일 PM 2:15 UTCinvestigating

   We are continuing to investigate this issue.

3. 2026년 6월 16일 PM 2:18 UTCidentified

   The team identified the issue is continuing to look into a fix

4. 2026년 6월 16일 PM 2:33 UTCidentified

   We are continuing to work on a fix for the issue

5. 2026년 6월 16일 PM 3:03 UTCidentified

   We have identified a fix and are working to implement

6. 2026년 6월 16일 PM 3:35 UTCidentified

   We have implemented the fix and are starting to see some recovery

7. 2026년 6월 16일 PM 3:49 UTCmonitoring

   A fix has been implemented and we are monitoring the results.

8. 2026년 6월 16일 PM 4:22 UTCresolved

   This incident has been resolved.

9. 2026년 6월 24일 PM 11:00 UTCpostmortem

   On June 16, 2026, an unintended configuration change introduced invalid data into a critical authentication service. As a result the service was unable to process authentication requests correctly, resulting in widespread authentication failures. This led to significant disruptions in order processing, reduced delivery volume, elevated cancellation rates, delayed deliveries, and increased customer remediation costs. The incident was resolved by restoring a known-good configuration, increasing service capacity to accelerate recovery, and temporarily managing incoming traffic while systems stabilized. To reduce the likelihood and impact of similar incidents in the future, we are implementing stronger configuration validation and schema enforcement, improving resilience and fallback behavior in critical services, and adding additional safeguards and rollout controls for changes to tier-0 infrastructure components.

1. 2026년 5월 5일 PM 3:47 UTCinvestigating

   We are currently investigating this issue.

2. 2026년 5월 5일 PM 4:50 UTCinvestigating

   We are continuing to investigate this issue.

3. 2026년 5월 5일 PM 6:12 UTCinvestigating

   We have found root cause and mitigated the issue.

4. 2026년 5월 5일 PM 6:14 UTCresolved

   This incident has been resolved.

5. 2026년 5월 7일 AM 12:38 UTCpostmortem

   ### **What happened** On May 5, 2026, a subset of partners integrated with DoorDash through openapi endpoints saw degraded connectivity. During the impacted window, some partner’s requests experienced elevated latency, errors, or timeouts. The affected merchants were unable to confirm orders within normal response windows. Our automated platform safeguard temporarily paused the stores on the impacted providers to help prevent additional failed order experiences. The issue was identified, mitigated, and fully resolved. Affected stores were brought back online once normal platform behavior was restored. ### **\[High Level\] Root cause**  The disruption was triggered when a partner inadvertently routed requests to an unintended endpoint with unexpected request payload. This resulted in degradation of a shared platform component responsible for handling API traffic. The degradation of this shared component led to elevated latency and timeouts for legitimate api requests. Interpreting these timeouts as store-level failures, our health-monitoring systems temporarily paused the affected stores to protect the overall ordering experience. ### **What we are improving** * Strengthening isolation between api traffic sources so that an issue from one path is less likely to affect another partner. * Conduct a comprehensive audit of platform paths and onboarding procedures to bolster overall system resilience. ### **Closing statement** We offer our sincere apologies to all merchants, partners, dashers, and consumers impacted by this disruption. DoorDash remains dedicated to providing a reliable platform for our integrated partners.  We are focused on utilizing the insights gained from this event to strengthen our recovery processes and platform stability.

1. 2026년 4월 15일 AM 8:33 UTCinvestigating

   We are currently investigating this issue.

2. 2026년 4월 15일 AM 8:46 UTCresolved

   This incident has been resolved.

1. 2026년 4월 15일 AM 8:30 UTCinvestigating

   Doordash order health seems to be impacted partially, seeing higher than expected latency. The team is actively investigating the issue.

2. 2026년 4월 15일 AM 8:45 UTCinvestigating

   The team has investigated and resolved the issue. Seeing api metrics back in trend.

3. 2026년 4월 15일 AM 8:46 UTCresolved

   This incident has been resolved.

1. 2026년 3월 19일 PM 4:22 UTCinvestigating

   We have detected an issue with Event Logs in Developer Portal. We are actively investigating.

2. 2026년 3월 19일 PM 4:49 UTCidentified

   The issue has been identified and a fix is being implemented.

3. 2026년 3월 19일 PM 5:57 UTCmonitoring

   A fix has been implemented and we are monitoring the results.

4. 2026년 3월 19일 PM 10:46 UTCresolved

   The incident has been resolved.

1. 2025년 12월 12일 PM 3:15 UTCidentified

   We are aware of users to be unable to access developer portal. The team is actively investigating the issue.

2. 2025년 12월 12일 PM 3:42 UTCmonitoring

   A fix has been deployed and we are monitoring auth errors.

3. 2025년 12월 12일 PM 4:14 UTCresolved

   We are no longer seeing the errors, Dev Portal should be available now.

1. 2025년 12월 8일 PM 12:47 UTCinvestigating

   Doordash order health seems to be impacted partially. The team is actively investigating the issue.

2. 2025년 12월 8일 PM 1:29 UTCresolved

   The order health has recovered and incident is now resolved.

3. 2025년 12월 10일 PM 5:11 UTCpostmortem

   Doordash experienced a partial service disruption affecting the core endpoints starting at approximately 3:50 AM PST on Dec 8. The issue was caused due to an elevated resource usage on our shared infrastructure, which led to connection problems that cascaded through multiple services, ultimately impacting several critical endpoints.

1. 2025년 12월 5일 AM 9:01 UTCinvestigating

   We are currently experiencing service degradation due to an outage with Cloudflare. This issue is causing platform-wide issues with DoorDash and may cause errors for some users. Our engineering teams are monitoring the situation and working to mitigate customer impact.

2. 2025년 12월 5일 AM 9:03 UTCinvestigating

   We are continuing to investigate this issue.

3. 2025년 12월 5일 AM 9:26 UTCmonitoring

   We are seeing recovery across our systems and are monitoring to confirm full restoration.

4. 2025년 12월 5일 AM 9:35 UTCresolved

   This incident has been resolved.

1. 2025년 12월 3일 PM 10:16 UTCinvestigating

   We have detected an issue with Drive Order Creation V1 API. We are actively investigating. API calls to the Create Delivery V1 endpoint that fail may be successful on retry. Next update in 15 minutes.

2. 2025년 12월 3일 PM 10:27 UTCmonitoring

   The issue with Order Creation is resolved but we will continue to monitor. Next update in 15 minutes.

3. 2025년 12월 3일 PM 10:45 UTCresolved

   The issue with Order Creation is now confirmed to be resolved. No further updates.

1. 2025년 12월 3일 AM 1:36 UTCinvestigating

   Partial order volume impact

2. 2025년 12월 3일 AM 1:38 UTCresolved

   This incident has been resolved.

1. 2025년 10월 20일 PM 2:30 UTCresolved

   We experienced an incident affecting certain DoorDash orders as a result of an AWS outage that impacted our partners. The incident has been fully mitigated.

1. 2025년 10월 17일 PM 7:36 UTCresolved

   From 9am to 12:49pm PT a new validation rule on DoorDash Drive was unexpectedly applied, requiring non-empty first and last names. This rule caused order creation requests with an empty customer first or last name to be rejected across API integrations and through the "Request a Delivery" form on the Merchant portal. After the rule change was rolled back by 12:49pm PT, the previous name validation rules were being applied and orders with empty first or last name were being accepted.

1. 2025년 10월 14일 PM 6:00 UTCresolved

   From 7:55 PM PDT on October 7, 2025 to 11:15PM PDT October 7, 2025, Drive experienced degraded ETA accuracy due to an internal database corruption issue. During this time, ETA estimates appeared significantly longer than expected for all Drive merchants. The underlying issue was partially resolved by 11:15 PM PDT and fully resolved by 2:25 AM PDT on October 8, 2025, and ETA accuracy has since returned to normal.

1. 2025년 9월 11일 PM 3:45 UTCinvestigating

   We have detected an issue with Order Creation. We are actively investigating.

2. 2025년 9월 11일 PM 3:50 UTCinvestigating

   We are continuing to investigate this issue.

3. 2025년 9월 11일 PM 3:51 UTCidentified

   The issue has been identified and working to be mitigated. We are seeing recovery

4. 2025년 9월 11일 PM 3:54 UTCmonitoring

   A fix has been implemented and we are monitoring the results.

5. 2025년 9월 11일 PM 4:02 UTCresolved

   This incident has been resolved.

1. 2025년 9월 9일 PM 7:03 UTCinvestigating

   We have detected an issue with Order Creation. We are actively investigating. Next update in 15 minutes.

2. 2025년 9월 9일 PM 7:07 UTCidentified

   We have identified an issue with Order Creation linked to one of our external vendors. The vendor is actively working on mitigation, and we are continuing to closely monitor the situation. Next update in 15 minutes.

3. 2025년 9월 9일 PM 7:13 UTCmonitoring

   The issue with Order Creation and Fulfillment is resolved but we will continue to monitor. Next update in 15 minutes.

4. 2025년 9월 9일 PM 7:31 UTCresolved

   The issue with Order Creation is now confirmed to be resolved. No further updates.

5. 2025년 9월 10일 PM 6:09 UTCpostmortem

   On September 4th, beginning around 11:30 AM PT, an external vendor \(Cloudflare\) experienced an outage only in Ashburn that caused partial degradation of Drive public-facing APIs. Impact was only for the requests around that region. Cloudflare applied a fix on their end, and by approximately 12:02 PM PT error rates returned to baseline. DoorDash eng teams remain engaged, worked closely with Cloudflare to resolve the issue and closely monitored recovery to ensure services remain stable.

1. 2025년 8월 6일 PM 8:49 UTCinvestigating

   Impact since 8/6 7am PST

2. 2025년 8월 6일 PM 9:04 UTCresolved

   Team identified cause and issue is now mitigated as of 2pm.

1. 2025년 8월 5일 PM 7:10 UTCinvestigating

   We are currently investigating this issue.

2. 2025년 8월 5일 PM 7:18 UTCinvestigating

   We have recovered and root causing the error spikes.

3. 2025년 8월 5일 PM 7:22 UTCmonitoring

   A fix has been implemented and we are monitoring the results.

4. 2025년 8월 5일 PM 7:27 UTCmonitoring

   We are continuing to monitor for any further issues.

5. 2025년 8월 5일 PM 7:28 UTCresolved

   This incident has been resolved. From 11:42 to 11:48, 12:02 to 12:10 we had two API 500 error spikes caused by internal DB outage. During the incident 7% of Drive Classic requests got impacted; and 25% of Drive API requests got impacted. Retries during the time would have worked if the clients have retry logic.

1. 2025년 8월 5일 PM 8:03 UTCresolved

   This incident has been resolved. From 11:42 AM to 11:48 AM and 12:02 PM to 12:10 PM PT, due to internal outage there was drop in order volume. We have since recovered.