On March 20, 2025, starting at 15:43 AM UTC, Zendesk users globally encountered 503 “Service Unavailable” errors and 5xx server-side issues, disrupting access to critical support tools and communication channels. While immediate mitigations stabilized core services, intermittent issues continued for over 24 hours, underscoring the complexity of multi-pod infrastructure failures.
Timeline of Events
The 503 Service Unavailable responses were not immediately recognized as the root cause of the disruption. This delay hindered Zendesk’s ability to fully understand the scope and impact of the outage in a timely manner and slowed the process of restoring service and assisting customers.
The Zendesk outage crippled workflows for thousands of businesses relying on the platform for customer support, sales, and internal collaboration. With critical processes disrupted, teams struggled to deliver timely and effective customer service. Below are some of the key consequences:
Many users encountered 5xx errors, which indicated problems on the server side. Users across industries—from retail to healthcare—were abruptly locked out of Zendesk portals. Support teams couldn’t view tickets, update cases, or access customer history.
Some services needed additional time to restart, leading to intermittent errors—often at the worst possible moments for businesses trying to handle customer inquiries. As a result, support agents wrestled with inconsistent access and were forced to pause or redo tasks.
Zendesk’s core support tools went offline for large sections of the outage, limiting response times and workflow coordination. Web widgets used on company websites for direct customer engagement also went down, frustrating users who expected immediate assistance or quick self-service options.
Even though Zendesk reported “majority of services” were restored by March 21, intermittent errors lingered for more than 24 hours. This may have forced businesses to switch to manual processes.
Catchpoint’s Internet Sonar flagged the outage at 15:22 AM UTC, 21 minutes before Zendesk’s internal alerts.
The Internet Sonar dashboard shows Zendesk’s outage affecting multiple global locations, with 100% downtime reported across several cities.
The scatterplot above from Internet Sonar visualizes the Zendesk outage, showing a surge in failed tests (red markers) starting around 15:22 AM UTC. The concentration of failures indicates a widespread service disruption.
The Zendesk outage underscores why real-time visibility, proactive monitoring, and a deep understanding of third-party dependencies are critical to maintaining Internet resilience.
Zendesk’s internal team took 21 minutes to correlate user reports with the 503 errors already flagged by Catchpoint’s Internet Sonar. While this might not seem like a long time, every minute of downtime means lost revenue, frustrated customers, and operational disruptions. The longer it takes to pinpoint the issue, the longer it takes to fix it.
Without immediate visibility into where and why a problem is occurring, IT teams waste precious time in war rooms and finger-pointing exercises, trying to determine if the issue is internal or caused by a third-party provider.
No organization can afford to operate without independent, proactive visibility into their digital ecosystem. That’s where Catchpoint’s suite of Internet Performance Monitoring (IPM) tools comes in.
Internet Sonar is a powerful tool that eliminates guesswork by providing real-time, independent Internet health data. Using Internet Sonar, you’ll know whenever a third party has an outage, where they have it, how long it's been going on, and whether or not it's likely to affect you.
Internet Sonar doesn’t just tell you when a site is down after people tweet about it. This means no finger-pointing, no war rooms, just straightforward, intelligent, and reliable Internet health information so you can get ahead of productivity or experience-impacting 3rd-party incidents.
The Zendesk outage also exposed vulnerabilities of multi-pod architectures, where failures in one pod cascaded into issues across multiple regions. While these architectures are designed for scalability and redundancy, they introduce complexities that can extend downtime when something goes wrong.
In this case, even after initial recovery, intermittent failures continued for over 24 hours, preventing full service restoration. For companies reliant on cloud-based applications like Zendesk, this reinforces the need for deep visibility into third-party infrastructure dependencies to understand:
Catchpoint’s Internet Stack Map can help with this by showing a live view of the health of your digital service and the services it depends on.
By automatically discovering third-party dependencies, it helps organizations understand the health of their digital ecosystem at a glance. When one component fails, it’s clearly highlighted, making root-cause analysis seamless.
Learn more about preventing outages from our guide, or test drive Catchpoint for yourself in our guided product tour.
On March 20, 2025, starting at 15:43 AM UTC, Zendesk users globally encountered 503 “Service Unavailable” errors and 5xx server-side issues, disrupting access to critical support tools and communication channels. While immediate mitigations stabilized core services, intermittent issues continued for over 24 hours, underscoring the complexity of multi-pod infrastructure failures.
Timeline of Events
The 503 Service Unavailable responses were not immediately recognized as the root cause of the disruption. This delay hindered Zendesk’s ability to fully understand the scope and impact of the outage in a timely manner and slowed the process of restoring service and assisting customers.
The Zendesk outage crippled workflows for thousands of businesses relying on the platform for customer support, sales, and internal collaboration. With critical processes disrupted, teams struggled to deliver timely and effective customer service. Below are some of the key consequences:
Many users encountered 5xx errors, which indicated problems on the server side. Users across industries—from retail to healthcare—were abruptly locked out of Zendesk portals. Support teams couldn’t view tickets, update cases, or access customer history.
Some services needed additional time to restart, leading to intermittent errors—often at the worst possible moments for businesses trying to handle customer inquiries. As a result, support agents wrestled with inconsistent access and were forced to pause or redo tasks.
Zendesk’s core support tools went offline for large sections of the outage, limiting response times and workflow coordination. Web widgets used on company websites for direct customer engagement also went down, frustrating users who expected immediate assistance or quick self-service options.
Even though Zendesk reported “majority of services” were restored by March 21, intermittent errors lingered for more than 24 hours. This may have forced businesses to switch to manual processes.
Catchpoint’s Internet Sonar flagged the outage at 15:22 AM UTC, 21 minutes before Zendesk’s internal alerts.
The Internet Sonar dashboard shows Zendesk’s outage affecting multiple global locations, with 100% downtime reported across several cities.
The scatterplot above from Internet Sonar visualizes the Zendesk outage, showing a surge in failed tests (red markers) starting around 15:22 AM UTC. The concentration of failures indicates a widespread service disruption.
The Zendesk outage underscores why real-time visibility, proactive monitoring, and a deep understanding of third-party dependencies are critical to maintaining Internet resilience.
Zendesk’s internal team took 21 minutes to correlate user reports with the 503 errors already flagged by Catchpoint’s Internet Sonar. While this might not seem like a long time, every minute of downtime means lost revenue, frustrated customers, and operational disruptions. The longer it takes to pinpoint the issue, the longer it takes to fix it.
Without immediate visibility into where and why a problem is occurring, IT teams waste precious time in war rooms and finger-pointing exercises, trying to determine if the issue is internal or caused by a third-party provider.
No organization can afford to operate without independent, proactive visibility into their digital ecosystem. That’s where Catchpoint’s suite of Internet Performance Monitoring (IPM) tools comes in.
Internet Sonar is a powerful tool that eliminates guesswork by providing real-time, independent Internet health data. Using Internet Sonar, you’ll know whenever a third party has an outage, where they have it, how long it's been going on, and whether or not it's likely to affect you.
Internet Sonar doesn’t just tell you when a site is down after people tweet about it. This means no finger-pointing, no war rooms, just straightforward, intelligent, and reliable Internet health information so you can get ahead of productivity or experience-impacting 3rd-party incidents.
The Zendesk outage also exposed vulnerabilities of multi-pod architectures, where failures in one pod cascaded into issues across multiple regions. While these architectures are designed for scalability and redundancy, they introduce complexities that can extend downtime when something goes wrong.
In this case, even after initial recovery, intermittent failures continued for over 24 hours, preventing full service restoration. For companies reliant on cloud-based applications like Zendesk, this reinforces the need for deep visibility into third-party infrastructure dependencies to understand:
Catchpoint’s Internet Stack Map can help with this by showing a live view of the health of your digital service and the services it depends on.
By automatically discovering third-party dependencies, it helps organizations understand the health of their digital ecosystem at a glance. When one component fails, it’s clearly highlighted, making root-cause analysis seamless.
Learn more about preventing outages from our guide, or test drive Catchpoint for yourself in our guided product tour.
This is some text inside of a div block.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。