Jumpcloud 장애 기록

1. 2026년 10월 23일 오후 3:55 UTCinvestigating

   현재 JumpCloud 사용자 또는 Admin Portal을 로드할 때 오류의 보고서를 조사하고 있습니다. 30 분 이내에 추가 정보가 업데이트됩니다.

2. 2026년 10월 23일 오후 4:00 UTCmonitoring

   제3자 서비스 관련 오류를 확인했습니다. 오류율이 감소하고 모니터링을 계속하고 있습니다.

3. 2026년 10월 23일 오후 4:11 UTCmonitoring

   우리는 더 많은 문제를 모니터링하고 있습니다.

4. 2026년 10월 23일 오후 4:24 UTCmonitoring

   약 15:20 및 15:50 UTC 10 월 23, 2026, 일부 고객은 JumpCloud Admin Portal 및 User Portal을 안정적으로로드 할 수 없습니다. 감염된 사용자는 오류 페이지 또는 느린, 불완전한 페이지 부하를 볼 수 있습니다. 원인은 우리의 제 3 자 콘텐츠 전달 및 보안 제공 업체 중 하나에 네트워크 중단이었다, 이는 그들의 동부 미국 위치의 일부를 통해 트래픽에 영향을 미치는. 네트워크가 안정화되면, 두 개의 포털에 액세스가 정상으로 돌아와서 문제가 완전히 해결되었는지 확인했습니다. 우리는 당신이 JumpCloud에 신뢰할 수있는 액세스를 의존하는 방법을 알고 조직을 관리하기 위해, 우리는 당신이 당신과 당신의 사용자에 기인 한 혼란에 대해 미안 해요. 중단이 공급자와 함께 시작될 때, 연결되는 유지는 우리의 책임입니다. 우리는 이미이 종류의 이벤트에 대한 장소에 중복을 가지고 있으며, 우리는 적극적으로 실패가 더 빨리 일어나고 미래에 더 자동적으로 발생합니다. 우리는이 사건의 전체 검토를 수행하고 Post-Incident Review를 자세한 타임 라인, 루트 원인 및 우리가 반복을 방지하는 데 걸리는 단계로 게시합니다.

5. 2026년 10월 23일 오후 6:25 UTCresolved

   이 사건은 해결되었습니다.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 10월 22일 오후 4:21 UTCinvestigating

   당사는 미국 지역의 고객에 대한 장치 인증서 발급 및 갱신에 영향을 미치는 문제를 확인했습니다. 약 9:26 AM MT (15:26 UTC). 이 시간 동안, 새로운 장치 인증서를 요청하는 장치는 하나를 얻을 수 없습니다. 이것은 주로 새로 등록 된 장치 및 인증서가 갱신에 의한 기기에 영향을 미칩니다. 기존 유효 인증서가있는 장치가 영향을받지 않으며 정상적으로 기능을 계속하지 않습니다.

2. 2026년 10월 22일 오후 4:44 UTCidentified

   문제는 식별되고 수정이 구현됩니다.

3. 2026년 10월 22일 오후 4:50 UTCmonitoring

   수정이 구현되어 결과를 모니터링하고 있습니다.

4. 2026년 10월 22일 오후 5:00 UTCresolved

   이 사건은 해결되었습니다.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 9월 25일 오후 5:27 UTCinvestigating

   우리는 현재 관리자 포털에서 명령에 대한 세부 사항이나 협회를 업데이트 한 성능의 평가 보고서를 조사하고 있습니다. 우리는 현재 문제의 원인을 조사하고 한 번 해결 된 상태 이벤트를 업데이트합니다.

2. 2026년 9월 25일 오후 5:29 UTCidentified

   문제는 식별되고 수정이 구현됩니다.

3. 2026년 9월 25일 오후 5:55 UTCmonitoring

   수정이 구현되어 결과를 모니터링하고 있습니다.

4. 2026년 9월 25일 오후 6:33 UTCresolved

   이 사건은 해결되었습니다.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 9월 25일 오후 5:22 UTCinvestigating

   우리는 현재 관리자 포털에서 명령에 대한 세부 사항이나 협회를 업데이트 한 성능의 평가 보고서를 조사하고 있습니다. 우리는 현재 문제의 원인을 조사하고 한 번 해결 된 상태 이벤트를 업데이트합니다.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 9월 14일 오후 1:33 UTCinvestigating

   현재 암호 재설정 후 문제가 발생하고 있습니다. RADIUS Sign-ins (VPN/Wi-Fi)는 새로운 암호로 실패할 수 있습니다. 그러나 - 영향받은 사용자는 여전히 오래된 암호로 로그인 할 수 있습니다. 우리는 현재 문제의 원인을 조사하고 1 시간 이내에 업데이트를 제공 할 것입니다.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 9월 14일 오후 1:25 UTCinvestigating

   현재 암호 재설정 후 문제가 발생하고 있습니다. RADIUS Sign-ins (VPN/Wi-Fi)는 새로운 암호로 실패할 수 있습니다. 그러나 - 영향받은 사용자는 여전히 오래된 암호로 로그인 할 수 있습니다. 우리는 현재 문제의 원인을 조사하고 1 시간 이내에 업데이트를 제공 할 것입니다.

2. 2026년 9월 14일 오후 1:58 UTCmonitoring

   수정이 구현되어 결과를 모니터링하고 있습니다.

3. 2026년 9월 14일 오후 2:38 UTCresolved

   이 사건은 이제 해결되었습니다.

4. 2026년 9월 21일 오후 8:49 UTCpostmortem

    #**Incident 보고서** **일시**: 9월 21, 2026 **일시:** 9월 14, 2026 **Description**: RADIUS 인증 실패에 대한 루트 원인 분석 EU 사용자 ** 등록: ** 유럽 연합 \ (EU \) ** 일정:** 9 월 14, 2026, JumpCloud는 유럽 연합 지역의 사용자를위한 RADIUS 인증에 영향을 미치는 문제를 식별했습니다. 최근 암호를 변경한 사용자들은 새로운 비밀번호로 로그인할 수 없었으며, 이전 암호는 계속 작동합니다. 이 문제는 계획 된 인프라 업그레이드를 따르는 기본 데이터베이스에서 업데이트를 중단 한 EU 지역의 데이터베이스 복제로 시작되었습니다. 결과적으로, 유럽 연합 (EU) 지역의 RADIUS 인증은 디렉토리 데이터의 최신 사본에서 제공되었으며, 최근 계정 변경 사항은 반영되지 않았습니다. JumpCloud는 유럽 연합 (EU) RADIUS 인증을 다른 지역에서 건강한 데이터베이스 복제하는 재조합 된 서비스를 복원 한 다음 영향을받는 EU 복제를 재구성하고 재조합합니다. EU 지역의 인증은 완전히 같은 날 복원되었습니다. **Happened:** 계획된 인프라 업그레이드의 일환으로 RADIUS 인증을 지원하는 데이터베이스 인스턴스가 배치되었습니다. 유럽 연합 (EU) 지역에서 데이터베이스 읽기 복제는 업그레이드 후 복제를 재개하지 않고 9 월 10, 2026의 기본 데이터베이스에서 업데이트를 중단했습니다. 다른 지역의 동등한 복제는 일반적으로 회복; EU 복제 만 영향을 받았다. 유럽 연합 (EU) 지역의 RADIUS 인증이 이 복제에서 제공되었기 때문에 현재 9 월 10 일까지 인 디렉토리 데이터 사본에 의존했습니다. 그 시점이 EU 지역에 반영되지 않았을 때 어떤 계정 변경이 해결되었습니다. 가장 눈에 띄는 symptom은 암호 변경이었습니다 - 새로 설정된 암호는 이전 암호가 여전히 작동하면서 실패합니다. 결과적으로 영향을받는 창 중 EU 지역에서 만든 계정 변경은 사용자의 액세스를 제거하거나 중단하기위한 변경 사항이 포함되지 않을 수 있습니다. 문제가 해결되고 복제가 재동식되면 모든 폐기 변화가 적용되고 EU 디렉토리 데이터가 완전히 현재 상태로 반환됩니다. ** 기여 인자:** 9 월 10과 9 월 14 사이의 감지 지연은 영향을받는 데이터베이스 인스턴스가 중단되었더라도 인프라 수준에서 건강한 상태를보고하기 위해 계속되었습니다. 우리의 경고는 이러한 인스턴스 수준의 가용성 메트릭 주위를 지향하고이 특정 복제 조건을 커버하지 않았다, 그래서 정지 복제는 자동화 된 알람을 트리거하지 않았다, underlying 문제. 이 탐지 간격을 닫는 것은 아래에 정확한 행동의 초점입니다. ** 정확한 행동:** JumpCloud는 recurrence를 방지하기 위해 다음 조치를 취합니다. 1.**Replication-health 모니터링 및 경고.** 우리는 데이터베이스 복제가 복제, 일반 가용성 상태의 독립적으로 중지 할 때 감지하는 자동화 된 경고를 추가하므로이 클래스의 실패가 즉시 캡처됩니다. 상태: 진행 중. 2. ** 인증 트래픽에 대한 엄격한 읽음 검증. ** 우리는 우리의 건강 검사를 강화하고 그래서 완전히 동기화되지 않은 복제는 서빙 인증 트래픽에서 자동으로 제거되고, 요청은 최신 데이터에서만 응답됩니다. 상태: 계획. 3. ** 인프라 변경에 대한 복제 유효성 검사. ** 우리는 데이터베이스 인스턴스 변경에 대한 우리의 절차를 업데이 트됩니다 그래서 그들은 다음에 진행하기 전에 완전히 복구 된 명시적 검증과 함께 한 지역에 적용된다. 상태: 진행 중.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 9월 3일 오전 6:07 UTCinvestigating

   JumpCloud는 현재 Dynamic Group에 사용자 및 시스템 할당 지연을 경험하고 있습니다. Administrators는 이러한 그룹 회원 업데이트에서 지연을 볼 수 있으며 연결 된 리소스에 대한 규정에 따라 지연됩니다.

2. 2026년 9월 3일 오전 6:14 UTCidentified

   문제는 식별되고 수정이 구현됩니다.

3. 2026년 9월 3일 오전 8:56 UTCidentified

   이 문제를 해결하기 위해 계속 노력하고 있습니다.

4. 2026년 9월 3일 오전 10:05 UTCmonitoring

   수정이 구현되어 결과를 모니터링하고 있습니다.

5. 2026년 9월 3일 오전 10:54 UTCmonitoring

   우리는 더 많은 문제를 모니터링하고 있습니다.

6. 2026년 9월 3일 오후 12:22 UTCresolved

   이 사건은 해결되었습니다.

공식 인시던트 업데이트를 자동 번역했습니다.

1. 2026년 6월 22일 오후 2:21 UTCinvestigating

   We are currently investigating delays or timeouts loading the user console. We are investigating the cause of the issues currently, and will provide an update within one hour.

2. 2026년 6월 22일 오후 3:08 UTCidentified

   We have identified that a third party outage is impacting our service. More information can be found at https://www.cloudflarestatus.com/incidents/v3yl7jqmqj51

3. 2026년 6월 22일 오후 4:39 UTCresolved

   The third party issue impacting our service has been resolved. Our internal monitoring is showing service states returned to normal.

4. 2026년 6월 25일 오후 8:00 UTCpostmortem

    ‌ **Date**: Jun 25, 2026 **Date of Incident:** Jun 22, 2026 **Description**: RCA for Third-Party CDN Network Disruption — Cloudflare / Zayo Fiber Cut ‌ #### **Summary:** On June 22, 2026, beginning at approximately 13:25 UTC, some JumpCloud customers experienced degraded access to web-facing services including the Admin Console, User Portal, API endpoints, and SSO authentication flows. Customers connecting through North America or accessing services routed through North American infrastructure were most affected. The root cause was a fiber cut in Eastern North America, a physical break in underground or undersea cables carrying internet traffic. Cloudflare attributed the disruption to Zayo, a network transit provider, experiencing an outage on some of its network routes, which caused reachability issues for services routing through those paths. Because JumpCloud's traffic transits the Cloudflare network before reaching JumpCloud's origin infrastructure, degradation at the Cloudflare/Zayo layer directly impacted our customers' ability to reach JumpCloud services. By late morning, most affected platforms had stabilized, with Cloudflare indicating its fix was being actively rolled out. JumpCloud's backend services and data plane remained fully operational throughout the event. The issue was isolated to the traffic path between end users and JumpCloud's origin infrastructure. ‌ #### **What Happened:** JumpCloud utilizes Cloudflare as the security and networking layer in front of all its global infrastructure. All HTTP/HTTPS requests to JumpCloud's web services traverse the Cloudflare network before reaching JumpCloud's origin infrastructure. ‌ #### **Provider Root Cause:** Cloudflare engineers first detected elevated error rates and latency at approximately 13:25 UTC on June 22, 2026. By 14:37 UTC, they had traced the root cause to a fiber cut in Eastern North America. Cloudflare confirmed that Zayo, a network provider, was experiencing an outage on some of its network routes, causing sites and services routing through those paths to become unreachable. Cloudflare's status page indicated that customers connecting through North America or accessing services in Europe may have seen increased latencies and timeouts as Cloudflare engineers worked to mitigate the issue. Traffic engineering efforts successfully mitigated the majority of congestion and packet drops, with services reported as largely stable with only minor residual impact remaining. Cloudflare's scheduled Newark \(EWR\) datacenter maintenance overlapped in timing but was a separate event - it was not the root cause of the fiber cut outage. ‌ #### **Corrective Actions \(Target End-of-July 2026\)**: While this incident was caused by a third-party provider we recognize the impact it had on our customers. JumpCloud is committed to reducing our exposure to single-provider failures and improving our resilience posture. The following corrective actions are actively underway: * **Multi-CDN Redundancy:** Cloudflare remains part of our architecture. We are adding a parallel, independently operated CDN path with weighted routing and security parity across both, so traffic can shift automatically if one path degrades. When complete, JumpCloud's edge will be capable of routing around a complete outage of any single CDN provider without customer action or service interruption. This work is in active delivery within our Platform Engineering organization. * **Resilient DNS Management:** We are modernizing how we manage DNS with phased, automated changes and rollback capabilities, reducing the risk that DNS operations themselves become a source of disruption. * **Validated Failover:** We are running controlled failover exercises including a production-scale test before relying on this redundancy in production. * **Enhanced Third-Party Monitoring & Automated Alerting:** We are expanding our monitoring to include automated detection of third-party provider degradation, enabling faster incident declaration and more rapid customer communication. This includes synthetic monitoring from multiple geographic regions that can differentiate between JumpCloud-origin issues and upstream provider issues ‌ #### **Provider Remediation:** Cloudflare has communicated the following remediation actions: * **Transit Path Restoration:** Cloudflare is coordinating with Zayo and fiber providers to repair the affected Eastern North American routes and restore full transit capacity. * **Traffic Engineering Response:** Cloudflare's traffic engineering teams implemented manual rerouting to redistribute load across unaffected network paths during the incident window. * **Ongoing Investigation:** Cloudflare has indicated a full post-incident review is underway to assess automated failover capabilities for large-scale transit provider failures. ‌ JumpCloud takes the reliability and availability of our platform seriously. We understand that our customers depend on JumpCloud for critical identity and device management operations, and any disruption, regardless of its origin, impacts their business. We are actively investing in infrastructure resilience to ensure that our dependency on any single third-party provider does not create an unacceptable risk to service availability.

1. 2026년 5월 8일 오전 6:41 UTCinvestigating

   We are currently investigating delays or timeouts loading the devices page in the admin console. We are investigating the cause of the issues currently, and will provide an update within one hour.

2. 2026년 5월 8일 오전 6:44 UTCidentified

   The issue has been identified and the team is working on the fix.

3. 2026년 5월 8일 오전 6:51 UTCmonitoring

   A fix has been implemented. We are currently monitoring the results.

4. 2026년 5월 8일 오전 8:28 UTCresolved

   This incident has been resolved.

1. 2026년 4월 28일 오전 8:01 UTCinvestigating

   We are currently aware of reports with agent Installation failing, this is affecting MacOS, Windows and Linux. We are investigating the cause of the issues currently, and will provide an update within one hour.

2. 2026년 4월 28일 오전 8:57 UTCidentified

   The issue has been identified and a fix will be implemented soon.

3. 2026년 4월 28일 오전 9:00 UTCmonitoring

   A fix has been implemented, and Agent installations should now be functioning as expected. Our team is actively monitoring the situation to ensure continued stability.

4. 2026년 4월 28일 오전 9:55 UTCresolved

   The incident has been resolved.

1. 2026년 4월 2일 오후 5:08 UTCinvestigating

   We are currently investigating a delay in LDAP directory synchronizations. While updates are processing, changes to user attributes and passwords may take longer than expected to propagate. The team is actively working to increase processing capacity and accelerate synchronizations.

2. 2026년 4월 2일 오후 5:16 UTCidentified

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

3. 2026년 4월 2일 오후 5:23 UTCmonitoring

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

4. 2026년 4월 2일 오후 5:55 UTCresolved

   This incident has been resolved.

1. 2026년 3월 31일 오전 12:32 UTCidentified

   JumpCloud is currently experiencing dispatch delays for core Directory services. Administrators may experience delays updating associations for users, groups, policies, directories, and commands. We have identified the cause of the issue and are actively implementing a fix.

2. 2026년 3월 31일 오전 2:38 UTCidentified

   We have made progress in reducing the backlog affecting Directory services and are continuing to work through the remaining queue. Administrators may still experience delays when updating associations for users, groups, policies, directories, and commands. The team is actively implementing additional changes to increase processing capacity and accelerate resolution. We will provide further updates in one hour.

3. 2026년 3월 31일 오전 3:42 UTCidentified

   The backlog affecting Directory services has been significantly reduced and continues to decrease. Administrators may still experience delays when updating associations for users, groups, policies, directories, and commands. The team continues to work on additional changes to accelerate processing. We will provide further updates in an hour.

4. 2026년 3월 31일 오전 5:00 UTCmonitoring

   The backlog affecting Directory services continues to decrease and we are approaching resolution. Administrators may still experience some delays when updating associations. The team is continuing to monitor, and our next update will be to confirm full resolution.

5. 2026년 3월 31일 오전 6:24 UTCresolved

   This incident has been resolved.

6. 2026년 4월 7일 오후 2:56 UTCpostmortem

    **Date**: Apr 7, 2026 **Date of Incident:** Mar 30, 2026 **Description**: RCA for Directory Association Processing Delays ‌ **Summary:** Starting March 30th at approximately 15:40 MDT, JumpCloud customers experienced significant delays in directory-related updates. This included latency in password changes, user-to-group associations, and outbound provisioning reflecting in downstream systems. The root cause was identified as a specific code deployment in our Devices service that inadvertently flooded a background processing queue with unpartitioned messages, causing a bottleneck that prevented updates from processing in real-time. The issue was fully resolved by 00:25 MDT on March 31, 2026. ‌ **What Happened:** The incident was caused by a change in how the JumpCloud agent retrieves software application configurations. 1. **Traffic Spike:** The new code shifted the "source of truth" for these configurations to a new database. If a device polled the system and did not find its record in the new database, the code automatically enqueued a "track collect" request to sync the data. 2. **Unexpected Volume:** We anticipated a "lazy backfill" \(where records are created over time\), but underestimated the number of devices that had no existing software bindings. This resulted in an immediate, massive spike of nearly 280,000 messages. 3. **The Bottleneck \(Partitioning\):** Crucially, these specific messages were enqueued without a "Partition ID." In our high-scale FIFO \(First-In-First-Out\) queue architecture, messages without a partition ID are processed one-by-one rather than in parallel. This effectively "serialized" the queue, preventing us from scaling up workers to process the backlog faster and causing the observed latency. ‌ **Resolution and Recovery**: Once the offending code was rolled back, the "tap" was turned off, and no further unpartitioned messages were added to the queue. Because the bottleneck was caused by the lack of partitioning, simply scaling horizontally could not speed up the processing of the existing backlog. The team monitored the queue throughput and determined that the safest and fastest path to recovery was allowing the worker to process the existing messages sequentially rather than risking further disruption by attempting to manually manipulate the production queue. ‌ **Corrective Actions**: To ensure this type of bottleneck does not occur again, we have committed to the following: * Improving pre-production testing to better simulate the scale and conditions that can occur in production queue processing * Reviewing other areas of the platform where similar patterns could produce unexpected request spikes * Enhancing monitoring and alerting thresholds to enable faster detection and response when queue backlogs begin to form * Strengthening our deployment validation process to more thoroughly account for background data migrations before releasing dependent code changes

1. 2026년 3월 12일 오후 4:54 UTCinvestigating

   We are currently investigating an issue with delays in syncing user, commands, and policy information with the JumpCloud Devices Agent backend service. We are investigating the cause of the issues currently, and will provide an update within one hour. Agent / Device logins are operating as expected.

2. 2026년 3월 12일 오후 7:24 UTCinvestigating

   We are continuing to investigate an issue with delays in syncing user, commands, and policy information with the JumpCloud Devices Agent backend service. Local Agent / Device logins without MFA or leveraging TOTP are operating as expected.

3. 2026년 3월 12일 오후 8:26 UTCidentified

   The issue has been identified and we have implemented a fix. We are starting to see some recovery with many agents checking in. We will update as our agent traffic normalizes.

4. 2026년 3월 12일 오후 9:08 UTCmonitoring

   Agent traffic is reaching normal levels, and the majority of customers are seeing full service restoration. Our engineering teams continue to actively monitor backend stability and traffic patterns. We expect to move to 'Resolved' status shortly as final systems stabilize.

5. 2026년 3월 12일 오후 9:48 UTCresolved

   This incident has been resolved.

6. 2026년 3월 17일 오후 9:25 UTCpostmortem

    ‌ **Date**: Mar 17, 2026 **Date of Incident:** Mar 12, 2026 **Description**: RCA for Agent Backend \(HAProxy\) System Degradation ‌ **Summary:** On March 12, 2026, from 10:05 AM to 2:45 PM MDT, JumpCloud experienced a significant service degradation affecting Agent-related activities. During this window, agent updates, including syncing users, passwords, policies and other agent data, as well as new agent installations were unavailable. This was caused by a "thundering herd" event triggered by a backend traffic-shaping change. We have since identified the root causes and implemented infrastructure changes to prevent a recurrence. ‌ **What Happened?** At 10:00 AM MDT, our engineering team enabled a feature flag \(a "circuit breaker"\) designed to protect our System Insights API from high load by returning `503 Service Unavailable` responses for certain non-critical requests. While the flag performed its intended function, it had an unforeseen secondary effect on the JumpCloud Agent’s connection logic. Because the agents could not reuse existing connections for these specific failed requests, hundreds of thousands of agents in our main production environment attempted to establish new mTLS \(mutual TLS\) connections simultaneously. This created a "Thundering Herd" event that saturated our HAProxy ingress layer, exhausting CPU resources and causing a cascade of connection failures. ‌ **Root Cause:** The prolonged nature of this incident was the result of three distinct, overlapping bottlenecks that our team had to isolate and resolve one by one: 1. **CPU-Intensive SSL Handshaking:** Establishing an mTLS connection is a CPU-intensive process. The sheer volume of simultaneous connection attempts pushed our HAProxy pods to their resource limits. This caused the pods to become unresponsive, leading to "Out of Memory" \(OOM\) kills and failed health probes. 2. **Health Check Death Spiral:** Our internal health checks initially relied on a Layer 7 SSL validation. Because the CPU was 100% occupied with agent reconnections, the pods couldn't respond to their own health checks in time. This caused the system to erroneously mark healthy pods as "down”, removing them from the rotation and further overwhelming the remaining pods. 3. **Load Balancer Handshake Saturation:** As we attempted to scale our infrastructure, the Application Load Balancer \(ALB\) encountered a throughput bottleneck specifically related to the rate of new connection establishments. The surge of agents attempting to negotiate new SSL handshakes at the same time exceeded the ALB's burst capacity, temporarily preventing even healthy backend pods from receiving and processing traffic. ‌ **Why It Took Time to Resolve:** While reverting the flag was the correct first step, the agents were already in an aggressive retry loop that continued even after the 503 errors stopped.  We had to experiment with several configurations \(adjusting health check intervals and timeout windows\) to find a balance that allowed pods to stay "alive" long enough to process the backlog. Stability was achieved only once we implemented Concurrency Control. By lowering the maximum allowed concurrent connections per pod, we stopped the CPU from over-committing to handshakes, allowing the system to reliably process a controlled flow of traffic until the global queue cleared. ‌ **Corrective Actions / Risk Mitigation:** **1.\) Edge Infrastructure Hardening** We are standardized on a new high-availability configuration for our HAProxy ingress layer. * **Concurrency Governance**: We have implemented a strict maxconn limit per pod. This acts as a "pressure valve," ensuring that the CPU remains available to process existing requests rather than becoming saturated by new connection attempts. * **Dynamic Capacity Management via Autoscaling**: We are implementing Horizontal Pod Autoscaling \(HPA\) for our HAProxy ingress layer, calibrated to trigger based on both CPU utilization and active connection counts.  This ensures we can absorb sudden traffic fluctuations and also maintain a controlled flow of requests to our backend services. **2.\) Agent Connectivity Optimization** We are updating the JumpCloud Agent’s communication layer to be more "network-aware" during degraded states: * **Enhanced Connection Pooling**: We are reconfiguring the agent's HTTP transport logic to maximize the reuse of existing idle connections. This significantly reduces the "Connection Tax" on our backend during high-traffic events. * **Streamlined Resource Handling**: We are implementing stricter protocols for draining and closing HTTP response bodies, ensuring that pooled connections are returned to the rotation immediately and reliably. **3.\) Adaptive Retry Logic \(Jitter\)** To further break up "synchronized" traffic spikes: * **Introduction of Jitte**r: While our agents currently use exponential backoff for poll requests, we are adding randomized "jitter" to our retry intervals. This spreads reconnection attempts across a wider window, preventing large blocks of agents from hitting the service at the exact same millisecond. * **Standardizing Resilient Retry Logic:** We are transitioning the Agent’s default HTTP client to a unified **exponential backoff** model for all request types. * **Controlled Rollou**t: This update will be managed via a staged rollout to monitor for any unforeseen side effects on fleet-wide connectivity patterns.

1. 2025년 11월 21일 오후 12:34 UTCinvestigating

   We had reports of delays or timeouts when accessing the Admin and User Portal consoles but now recovering. This issue has been traced to an ongoing network problem within the Asia region reported by Cloudflare. We are closely monitoring the situation and will provide an update as soon as new information becomes available.

2. 2025년 11월 21일 오후 12:44 UTCmonitoring

   Per the latest update from CloudFlare, a fix has been implemented and results are being monitored. https://www.cloudflarestatus.com/incidents/5p8c3l54b5cl

3. 2025년 11월 21일 오후 2:30 UTCresolved

   This incident has been resolved.

1. 2025년 11월 19일 오전 6:46 UTCinvestigating

   We're currently investigating reports of intermittent access issues with JumpCloud's Console Service for administrator at https://console.jumpcloud.com/. We are investigating the cause of the issues currently, and will update the status event once resolved.

2. 2025년 11월 19일 오전 7:00 UTCmonitoring

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

3. 2025년 11월 19일 오전 7:16 UTCresolved

   Issue is resolved and closed.

4. 2025년 11월 21일 오후 6:25 UTCpostmortem

    ‌ **Date**: Nov 21, 2025 **Date of Incident:** Nov 19, 2025 **Description**: RCA for Admin Portal Login Errors ‌ **Summary:** On November 19, 2025, starting at approximately 04:30 UTC, between 1-5% of requests experienced intermittent failures to successfully authenticate to the Admin Console, lasting until roughly 06:30 UTC.  Users attempting to authenticate received an “unexpected” error message during this window, but subsequent retries may have been successful. **Root Cause:** This issue was triggered during a standard infrastructure update and traffic shift intended to move services to a new, updated cluster.  The core issue was a combination of an infrastructure configuration mismatch and gaps in our detection and validation processes. 1. Configuration Drift: The new infrastructure cluster \(Green Cluster\), intended to host the service, was missing a single but essential configuration value used by the control plane’s service mesh. This value had been recently applied to the existing cluster \(old cluster\) but was inadvertently excluded when the new cluster's baseline configuration was created and branched.  When production traffic began routing to the new cluster, the missing configuration caused some access components to fail, leading to the login errors. 2. Detection Gaps: The application logged the configuration failure as a _Warning_ message, rather than a critical error. This meant our automated monitoring system did not trigger an immediate alert or rollback when the issue first occurred. The team quickly isolated the issue to the new Green Cluster, and an emergency process was initiated to immediately revert all production traffic back to the stable old cluster. ‌ **Corrective Actions / Risk Mitigation:** 1. Automated Configuration Diff Check - Implementing an automated process to continuously compare and ensure 100% configuration parity between old and new production clusters during all transition phases. 2. Clear Rule Enforcement - Reinforcing and automating the process to ensure all configuration changes are applied consistently across all active and future clusters. 3. Multi-Layer Error Monitoring - Implementing error rate monitoring at every layer of the network and application stack to ensure no failure goes undetected.

1. 2025년 11월 11일 오전 8:17 UTCinvestigating

   We are currently investigating reports of intermittent issues with launching Remote Assist tool. We are investigating the cause of the issue currently, and will provide an update within one hour.

2. 2025년 11월 11일 오전 8:40 UTCidentified

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

3. 2025년 11월 11일 오전 9:12 UTCmonitoring

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

4. 2025년 11월 11일 오전 9:41 UTCresolved

   This incident has been resolved.

1. 2025년 11월 6일 오후 12:28 UTCinvestigating

   We are currently investigating an issue with the authentication on SSO OIDC applications. We have already identified the cause and will provide an update within one hour.

2. 2025년 11월 6일 오후 12:33 UTCidentified

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

3. 2025년 11월 6일 오후 1:10 UTCmonitoring

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

4. 2025년 11월 6일 오후 1:47 UTCresolved

   This incident has been resolved.

5. 2025년 11월 13일 오후 11:33 UTCpostmortem

    ‌ **Date**: Nov 13, 2025 **Date of Incident:** Nov 6, 2025 **Description**: RCA for SSO/OIDC Service Degradation ‌ **Summary:** On November 6, 2025, starting at approximately 12:00 UTC, customers experienced failures to launch any application relying on JumpCloud's OIDC-based Single Sign-On \(SSO\), lasting for roughly one hour. ‌ **Root Cause:** The outage was caused by a combination of two errors during a scheduled compliance procedure: 1. Faulty Password Generation: Our automated system for rotating database passwords created a new credential that contained unsafe special characters. 2. Missing Special Character Logic: The entrypoint script for our core SSO service was missing logic to handle these special characters before using the password to construct a database connection string. When the SSO service attempted to restart and use the newly rotated password, the presence of the unsafe characters caused the connection string to be misinterpreted as invalid, leading to a parsing failure and service degradation. This issue stemmed from a latent configuration bug that was masked by prior rotation processes. Previously, database passwords were rotated manually using an older system \(`random_password` IAC resource\) which was explicitly configured to only generate alphanumeric characters. These characters are inherently safe in a URL context, so the underlying bug in the SSO service's connection logic was never exposed.  When the credential management was successfully migrated to the new, more robust rotation process, the new function began generating highly complex passwords, including special characters, for the first time. This immediately triggered the latent parsing flaw in the SSO service’s entrypoint script. ‌ **Corrective Actions / Risk Mitigation:** 1. Hardening code logic - All services that construct database connection strings will be audited and updated to explicitly encode the password component eliminating character misinterpretation. 2. Enhanced rotation alerting -  New monitoring and alerting dashboards are in place to track the health and success schedule of all automated credential rotation jobs, providing an immediate alert if a rotation creates an invalid credential. 3. Update password generation logic - The automated credential rotation function has been updated to explicitly generate passwords that are safe, avoiding complex, reserved characters.

1. 2025년 11월 4일 오전 11:40 UTCinvestigating

   We are seeing issues with SSO authentication. We are investigating this currently and will update within 1 hour

2. 2025년 11월 4일 오전 11:47 UTCinvestigating

   We are continuing to investigate this issue.

3. 2025년 11월 4일 오후 12:20 UTCidentified

   We have identified an issue that is causing intermittent login issues with the JumpCloud User Portal and Admin Portal. We have also identified issues with JumpCloud MFA, LDAP, RADIUS, and authentication with SSO. We are working on implementing a fix and will provide another update as soon as possible.

4. 2025년 11월 4일 오후 12:59 UTCidentified

   We continue to see intermittent issues with accessing the JumpCloud User and Admin Portal, MFA, LDAP, RADIUS, and SSO. During this time access attempts to LDAP and RADIUS are only impacted if a user is authenticating with MFA. Our team is working on implementing a fix and will provide another update as quickly as possible.

5. 2025년 11월 4일 오후 1:17 UTCmonitoring

   We have implemented a fix and users should now be able to access the User and Admin Console, MFA, LDAP, RADIUS, and SSO without issue.  We will continue to monitor the results of the fix.

6. 2025년 11월 4일 오후 2:19 UTCresolved

   Services have been fully restored and this incident has been resolved. We will provide a formal postmortem as a follow up.

7. 2025년 11월 7일 오후 5:47 UTCpostmortem

    ‌ **Date**: Nov 7, 2025 **Date of Incident:** Nov 4, 2025 **Description**: RCA for Auth Database Degradation ‌ **Summary:** On November 4, 2025, a number of customers experienced intermittent failures, timeouts and increased latency when attempting to authenticate to multiple JumpCloud Services, including consoles, LDAP, RADIUS and SAML, or use Multi-Factor Authentication. ‌ **Root Cause:** The incident was triggered by an issue in the deployment process involving a database schema change and a subsequent application code release. During this deployment, a planned database change unintentionally removed several database indexes required by the existing application code. The sequence of failure was as follows: 1. Deployment Order Error: The database schema change \(which removed necessary indexes\) was applied to the production database before the new application code \(which did not require those indexes\) was deployed. 2. Performance Collapse: The existing, high-volume authentication code \(used for functions like TOTP and push authentication\) was forced to run against the now-inefficient database structure. Queries that normally took milliseconds suddenly took several seconds. 3. Connection Exhaustion: These slow queries held database connections open for extended periods, quickly overwhelming the database server's available connection pool. 4. Full Outage: With no available connections, the main authentication API could not communicate with the database, leading to 100% CPU utilization on the database server and triggering the intermittent timeouts and failures experienced by our customers. ‌ **Why Testing Did Not Catch This:** The issue was not identified during testing in our Development or Staging environments due to insufficient Load Simulation. The resource consumption issues and connection exhaustion only manifest under the extreme pressure of peak production traffic volume. The simulated load profiles in our lower environments were not sufficient to expose this specific failure mode. ‌ **Corrective Actions / Risk Mitigation:** 1. Mandatory schema change review - All database schema changes must now undergo an additional level of review to explicitly assess index dependencies and impact. 2. New deployment phasing - We are implementing new tools and checks to enforce that application code dependent on a schema change is deployed before a database change is executed. 3. Enhance alerting -  We are implementing new monitors and alerts specifically for the Auth-API's database connection pool health and CPU utilization. 4. Enhanced load testing - We are revisiting the load profiles used in our staging environments looking for opportunities to more accurately simulate peak production traffic.

1. 2025년 10월 20일 오후 3:44 UTCinvestigating

   Due to an issue with our Cloud Provider, JumpCloud is experiencing intermittent issues with our PAM service and some Support tools which may affect case creation. We are investigating this with our provider and will update this incident as we know more.

2. 2025년 10월 20일 오후 3:50 UTCidentified

   We are monitoring our services and continue to work with our provider to restore full functionality.

3. 2025년 10월 20일 오후 5:43 UTCmonitoring

   We've seen an increase in recovery and are monitoring

4. 2025년 10월 20일 오후 9:46 UTCresolved

   This incident has been resolved.

5. 2025년 10월 27일 오후 7:21 UTCpostmortem

    **Date**: Oct 27, 2025 **Date of Incident:** Oct 20, 2025 **Description**: RCA for Service Degradation Linked to AWS US-EAST-1 Regional Disruption ‌ **Summary:** On October 20, 2025, between approximately 07:00 UTC and 10:00 UTC, the JumpCloud platform experienced significant performance degradation. This primarily affected the responsiveness of our core APIs, administrative console access, user console access, single sign-on \(SSO\) functions for customers, failed trial creation, user updates, and the potential loss of some Directory Insights events.  Residual performance degradation, affecting services like our Privileged Access Management \(PAM\) feature and specific Support Portal APIs, persisted until approximately 17:00 UTC, at which point all services were fully restored. ‌ The degradation was not caused by a failure within the JumpCloud platform code or infrastructure configuration, but was a direct consequence of a severe, cascading failure within the Amazon Web Services \(AWS\) US-EAST-1 region. Although many of our services are deployed across multiple Availability Zones \(AZs\) within the region for resilience, the nature of the AWS issue - impacting fundamental regional services - compromised inter-AZ communication preventing some of our standard failover mechanisms from operating successfully. Services returned to full operational status after AWS reported stability with the impacted foundational services. ‌ **Root Cause:** Based on the [post-incident analysis of AWS,](https://aws.amazon.com/message/101925/) the service disruption was not a single event but a sequence of cascading failures across three fundamental AWS services. ‌ 1. DynamoDB Failure Due to Latent DNS Race Condition \(Initial Trigger\) 2. EC2 Launch Congestion and Network Propagation Delays \(Sustained Impact\) 3. Network Load Balancer \(NLB\) Health Check Instability \(Final Phase\) ‌ All affected service teams were paged by our monitoring and alerting systems, and our incident management team was engaged to coordinate efforts and assess areas where we could throttle traffic to stabilize remaining capacity. ‌ ‌ **Corrective Actions / Risk Mitigation:** 1. While our current architecture mitigates some single-Availability Zone \(AZ\) failures, our primary focus is to eliminate single-region dependency entirely. JumpCloud is actively engaged in strategic engineering initiatives to further strengthen the platform's foundation. We are conducting a thorough review of cross-region dependencies and replication strategies to enhance our service's resilience against widespread environmental disruptions, always striving to meet the highest standards of availability.

1. 2025년 10월 20일 오전 7:52 UTCinvestigating

   Due to an issue with our Cloud Provider, JumpCloud is experiencing intermittent issues with multiple services. We are investigating this with our provider and will update this incident as we know more.

2. 2025년 10월 20일 오전 9:11 UTCidentified

   We are monitoring our services, and starting to see some recovery.

3. 2025년 10월 20일 오전 10:02 UTCmonitoring

   We are continuing to monitor as we see further recovery with all services.

4. 2025년 10월 20일 오전 10:14 UTCresolved

   This incident has been resolved.

5. 2025년 10월 27일 오후 7:18 UTCpostmortem

    **Date**: Oct 27, 2025 **Date of Incident:** Oct 20, 2025 **Description**: RCA for Service Degradation Linked to AWS US-EAST-1 Regional Disruption ‌ **Summary:** On October 20, 2025, between approximately 07:00 UTC and 10:00 UTC, the JumpCloud platform experienced significant performance degradation. This primarily affected the responsiveness of our core APIs, administrative console access, user console access, single sign-on \(SSO\) functions for customers, failed trial creation, user updates, and the potential loss of some Directory Insights events.  Residual performance degradation, affecting services like our Privileged Access Management \(PAM\) feature and specific Support Portal APIs, persisted until approximately 17:00 UTC, at which point all services were fully restored. ‌ The degradation was not caused by a failure within the JumpCloud platform code or infrastructure configuration, but was a direct consequence of a severe, cascading failure within the Amazon Web Services \(AWS\) US-EAST-1 region. Although many of our services are deployed across multiple Availability Zones \(AZs\) within the region for resilience, the nature of the AWS issue - impacting fundamental regional services - compromised inter-AZ communication preventing some of our standard failover mechanisms from operating successfully. Services returned to full operational status after AWS reported stability with the impacted foundational services. ‌ **Root Cause:** Based on the [post-incident analysis of AWS,](https://aws.amazon.com/message/101925/) the service disruption was not a single event but a sequence of cascading failures across three fundamental AWS services. ‌ 1. DynamoDB Failure Due to Latent DNS Race Condition \(Initial Trigger\) 2. EC2 Launch Congestion and Network Propagation Delays \(Sustained Impact\) 3. Network Load Balancer \(NLB\) Health Check Instability \(Final Phase\) ‌ All affected service teams were paged by our monitoring and alerting systems, and our incident management team was engaged to coordinate efforts and assess areas where we could throttle traffic to stabilize remaining capacity. ‌ **Corrective Actions / Risk Mitigation:** 1. While our current architecture mitigates some single-Availability Zone \(AZ\) failures, our primary focus is to eliminate single-region dependency entirely. JumpCloud is actively engaged in strategic engineering initiatives to further strengthen the platform's foundation. We are conducting a thorough review of cross-region dependencies and replication strategies to enhance our service's resilience against widespread environmental disruptions, always striving to meet the highest standards of availability.