All systems operational

Between 13:30 and 13:39 UTC on May 26, 2026, the [Files.com](http://Files.com) platform experienced a 9-minute issue during which uploads, as well as create, update, and delete operations against any resource, failed. Logins, downloads, and listings were unaffected during this period. Authentication-only and read-only workflows continued to operate normally.  The issue was caused by a routine maintenance event on our underlying Aurora MySQL database cluster, which moved the writer role from one database instance to another. Our application failed to recognize this change and continued attempting to write to the previous instance, which was now operating in a read-only capacity. We restored full write capability at 13:39 UTC by restarting the application, which forced it to reconnect to the correct database.  [Files.com](http://Files.com) is aware of this role changing mechanism and had built handling for it that had been previously tested in production. That code regressed during a framework upgrade and stopped functioning, and we failed to detect the regression because we did not have a standing test that exercises a database failover in our staging environment. We are correcting both gaps. We are restoring proper failover behavior in the application, and we are adding an automated failover test in staging so that this class of regression cannot ship undetected again.  We also identified that the maintenance window for our database cluster is currently scheduled during US business hours, which means routine, automated maintenance events occur at exactly the wrong time of day for our customers. We are moving this window to an off-peak time.  We promise a system that works perfectly, all of the time, and today we failed to deliver that to you. We are particularly disappointed because [Files.com](http://Files.com) had previously solved this exact failure mode and allowed a regression to ship. Our entire engineering team is working hard to prevent issues like this one from occurring in the future. If you need additional assistance or continue to experience issues, please contact our Customer Support team.

Started May 26, 2026, 02:06 PM UTC · Resolved May 26, 2026, 02:06 PM UTC

From 17:32 UTC until 19:14 UTC on April 1, 2026, a subset of servers in our primary US region returned TLS errors that produced elevated failure rates across the [Files.com](http://Files.com) Web UI, API, FTP/S, SFTP, and WebDAV. Customers whose traffic was routed to unaffected hosts experienced no issues; others saw login timeouts or "invalid password" messages. **What Happened** While adding an additional domain to our primary wildcard SSL certificate, a bug in our internal certificate management software generated a certificate that omitted the wildcard entries for many of our domains. Our automated certificate rotation process then installed that faulty certificate on a portion of the fleet, causing those hosts to reject connections. Because traffic is load-balanced evenly, some customer sessions failed while others succeeded, making the issue difficult to detect on global dashboards. We identified the certificate issue and generated a corrected certificate within 10 minutes. However, the incorrect certificate remained deployed for an additional 92 minutes. This second delay is the most frustrating part of this incident. Recent modifications to our certificate rotation system had not been reflected in our internal documentation. As a result, our on-call engineers were working from outdated instructions for manually forcing a certificate rotation. It took additional time to uncover the correct process for performing the manual rotation. **What We Have Done To Mitigate This In The Future** We have updated our certificate generation to avoid the accidental omission of wildcard entries through additional validation. We have added an additional monitoring check that which continuously verifies deployed certificates on every public endpoint for the correct wildcard entries. We have documented the certificate-rotation mechanism and emergency override procedure in our internal documentation. We know our customers rely on [Files.com](http://Files.com) for mission-critical workflows, and any service interruption is unacceptable. We apologize for the disruption and appreciate your patience as we improve our safeguards to ensure this does not recur.

Started Apr 1, 2026, 05:30 PM UTC · Resolved Apr 1, 2026, 05:30 PM UTC

From approximately 18:20 UTC to 19:23 UTC on 10 February 2026, customers could not preview or edit Office documents in the [Files.com](http://Files.com) Editor.  All other [Files.com](http://Files.com) services remained fully available.   **What happened** Our document-editing service relies on an internal message broker \(Amazon MQ / RabbitMQ\).  A routine broker maintenance reboot caused our editor software to stop accepting new editing sessions but continued to report itself as healthy.  Because our health check logic was flawed and yielding that incorrect status, our load balancer kept sending traffic to the failing servers, resulting in the spinning “Loading…” message you observed.   **What we found** * The editor’s own health endpoint did accurately represent its real status after the broker reboot but that was not known because the failure response was a HTTP 200 OK with the body of “false”. * Our Consul monitoring treated any HTTP 200 as healthy, so no automatic alert fired. * This same chain of events occurred on 20 January but was not fully remediated.   **What we are doing** 1. Updating the Consul health check to validate the response body and fail closed when the editor reports “false”. 2. Enabling CloudWatch logs and metrics for Amazon MQ and alerting on broker restarts and AMQP channel errors. 3. Adding an integration test that continuously opens and saves a document in production and pages engineering if it stalls. 4. We are working to replicate the exact error scenario in staging by creating a full production-like cluster. Once replicated, we will use that data to craft a solution to prevent this error from reoccurring.   We failed to detect the problem promptly and allowed it to recur.  We know you rely on the [Files.com](http://Files.com) Editor for time-critical collaborative work, and we are sorry for the disruption.  The actions above are already in progress, and we will publish further updates on our status page as milestones are reached.  Thank you for your continued trust in [Files.com](http://Files.com).

Started Feb 10, 2026, 06:30 PM UTC · Resolved Feb 10, 2026, 06:30 PM UTC

We have resolved an incident causing failures to load the Files.com Editor in our web interface. This incident was resolved at 19:45 UTC on January 20th. Office documents can once again be previewed and edited in the built-in Files.com Editor. We apologize for the inconvenience that this incident caused. We will perform a full postmortem on this incident and publish the report here when it is available.

Started Jan 20, 2026, 07:44 PM UTC · Resolved Jan 20, 2026, 08:02 PM UTC

From 21:45 UTC to 22:55 UTC on 9 January 2026, a subset of customers with custom domains enabled experienced elevated “500 – Internal Server Error” responses when uploading \(and in some cases downloading\) files through certain Remote Server Mount–backed transfer paths, including [Files.com](http://Files.com) Agent- backed connections. A code change in the [Files.com](http://Files.com) API control plane, intended to correct a bug in custom-domain uploads, introduced an unintended error condition affecting uploads routed through remote server mounts for sites with custom domains. Specifically, the API returned an invalid \(blank\) upload URL in its response, which caused subsequent upload steps to fail and resulted in HTTP 500 errors. These errors were immediately generated and logged by our application. Under normal conditions, this error volume would have triggered an automated PagerDuty alert to our on-call engineering team. However, a misconfiguration in our error-tracking system \(Sentry\) caused the majority of these error events to be suppressed. As a result, the expected alerts were not generated, and the issue was not immediately paged to on-call staff. Although our CI and production canary testing do exercise Agent-backed and Remote Server Mount upload workflows, those tests are not currently executed against customer sites using custom domains. Because this defect only manifested when custom domains were combined with remote server mounts, this edge case was not detected during pre-production testing. The issue was ultimately identified through customer reports, which prompted immediate investigation and remediation. A patch was verified in staging at 22:48 UTC and deployed to production at 22:55 UTC, fully restoring normal operation. We corrected the Sentry configuration and deployed new, version-controlled Terraform configuration to ensure that critical production errors always generate complete telemetry and alerts. We will also expand automated tests and deployment canaries to explicitly cover custom domain configurations so that this class of issue is detected during pre- production testing. This incident represents our most significant service disruption in the past year. While the issue affected a relatively small portion of our customer base, it had a material impact on those customers, with disruptions lasting more than an hour for some workflows. We are acutely aware that for the customers affected, the scope of the impact matters far more than the size of the cohort. We are extremely embarrassed by this failure. At the same time, we believe it is important to provide full transparency: even with this incident, [Files.com](http://Files.com) continued to meet and exceed its contractual SLA commitments for the month. That fact does not lessen the seriousness of the disruption, but it does highlight the overall resilience of the platform and our ongoing investment in reliability. We recognize that many customers depend on Remote Server Mounts for time- critical and business-critical workflows, and that any interruption is unacceptable. We sincerely apologize for the disruption this incident caused and for the impact to your operations. We take this failure seriously and are committed to maintaining the level of reliability and trust you expect from [Files.com](http://Files.com). If you have any questions or would like to discuss this incident further, our Support team stands ready to assist.

Started Jan 9, 2026, 04:00 PM UTC · Resolved Jan 9, 2026, 11:45 PM UTC

We have resolved an issue which caused elevated error rates in outbound connections from Files.com to Sharepoint and OneDrive remote servers. The impact was limited to these two remote server integrations and did not impact any inbound connections to Files.com that don’t involve Onedrive or SharePoint remote servers. This incident occurred between the times of 1:23PM and 2:58PM UTC.

Started Nov 25, 2025, 01:23 PM UTC · Resolved Nov 25, 2025, 01:23 PM UTC

From 10:03 UTC to 19:02 UTC on August 4, 2025, some users attempting to sign in to [Files.com](http://Files.com) via URLs ending in /login \(e.g., [acme.files.com/login](http://acme.files.com/login)\) encountered a red warning stating “Deceptive site ahead.” This warning appeared in Chrome, Safari, Firefox, and other browsers that rely on Google Safe Browsing. Affected access paths included: * Bookmarked links to /login * Direct navigation to /login * The "Sign In" button on our marketing website All other product surfaces—including APIs, desktop and CLI apps, agents, mobile apps, custom domains, and already-loaded web app sessions—remained fully functional and unaffected. **Root Cause** The Chrome, Safari, and Firefox browsers all use a service from Google called Google Safe Browsing to identify potentially dangerous sites and apply a full page warning when a user tries to visit any site that has been flagged by Google Safe Browsing. The [Files.com](http://Files.com) subdomain for a [Files.com](http://Files.com) customer was incorrectly reported to Google Safe Browsing as a phishing site by a third-party security vendor affiliated with that customer. Google accepted the report without independent validation and—unexpectedly— applied a warning to any login page on the entire [files.com](http://files.com) domain, not just that specific customer subdomain. To be clear: * No malware was present. * No compromise occurred. * No policy violation took place on the [Files.com](http://Files.com) platform. We have identified both the customer and the third-party vendor responsible for the report and have taken action to prevent any recurrence. **Response Timeline** At 10:09 UTC, our team convened an incident bridge within six minutes of detection and posted a status page update. Shortly after: * We disabled the affected customer’s subdomain. * We submitted a formal appeal to Google via Search Console, providing logs and proof that the report was a false positive. * We pushed a change to reroute new login sessions from /login to /newsession, restoring access for all customers except those with bookmarks to the previous URL. * We requested “This site is safe” confirmations from affected customers to help expedite Google’s manual review. At 19:02 UTC Google completed its review and removed the incorrect classification. Browser warnings began disappearing shortly thereafter as Safe Browsing lists refreshed \(typically within 30–60 minutes\). **Our Perspective** We understand the critical nature of [Files.com](http://Files.com) in your workflows, and we consider this incident wholly unacceptable—not because of any fault in our systems, but because access to our login page was unfairly blocked, creating real-world disruption for our customers and reputational damage for both you and us. We are deeply frustrated by how this situation was handled by Google: * An erroneous third-party report—submitted via a poorly governed “security” company system—was accepted by Google without any review. * Worse, the impact of that report was applied far beyond its scope, extending the warning to our entire domain, rather than the single customer subdomain in question. * Despite escalating through multiple internal and public channels, it took Google over 9 hours to act, during which time countless users encountered misleading and damaging warnings when trying to access [Files.com](http://Files.com). This incident reveals a deeply troubling reality: Google Safe Browsing wields extraordinary power over the web, but operates with little transparency, no meaningful appeals process, and effectively zero accountability. There’s no SLA, no phone number, and no obligation to fix their own errors promptly—if at all. We believe this system is broken. A single bad report—no matter how inaccurate— can cause cascading damage to legitimate businesses without recourse. That’s not just frustrating. It’s dangerous. We’re currently working to establish a more reliable escalation path within Google Safe Browsing, but the truth is: no business should have to navigate this kind of opaque and unilateral system to protect its integrity. We sincerely apologize for the disruption and appreciate your patience and trust. If you have any further questions or concerns, please reach out to our Support team.

Started Aug 4, 2025, 12:28 PM UTC · Resolved Aug 4, 2025, 09:09 PM UTC