Outage recap · DeepL

DeepL outage — February 18, 2025

On February 18, 2025, DeepL had 3 reported incidents on DeepL's official status page. Below is what was reported — impact, timing, and the AI services the same upstream outage could affect.

Elevated Error Rate on DeepL SSO

minor
Started
Feb 18, 6:32 AM EST
Resolved
Feb 18, 7:29 AM EST
Duration
57 min

DeepL's updates (6)

  1. investigating · Feb 18, 6:32 AM EST

    We are seeing elevated error rates in our login system and investigate the issue.

  2. investigating · Feb 18, 7:11 AM EST

    We are rolling back a change that was done to our login system.

  3. monitoring · Feb 18, 7:13 AM EST

    Error rate is back to normal after the rollback. We are monitoring the situation.

  4. monitoring · Feb 18, 7:25 AM EST

    We keep monitoring after a short return of errors.

  5. resolved · Feb 18, 7:29 AM EST

    The incident has been resolved. DeepL SSO login functionality is restored. We apologize for the inconvenience.

  6. postmortem · Feb 18, 8:34 AM EST

    We have migrated our SSO service to a new system to improve performance. During this process, some users experienced problems logging in due to outdated connection records being cached. We quickly resolved the issue by updating our systems to support both the old and new configurations, ensuring that all users were able to access the service. To prevent this from happening again, we will improve the way we handle system changes and monitor connection updates more closely. We apologize for any inconvenience and thank you for your patience.

Updates are DeepL's own words from its status page (status: postmortem); we never add an inferred cause. The vendor's incident page ↗

Partial Outage on Write Pro

major
Started
Feb 18, 8:19 AM EST
Resolved
Feb 18, 8:33 AM EST
Duration
14 min

DeepL's updates (7)

  1. investigating · Feb 18, 8:19 AM EST

    We are currently seeing errors for a subset of languages on Write Pro and are investigating.

  2. investigating · Feb 18, 8:19 AM EST

    We are continuing to investigate this issue.

  3. investigating · Feb 18, 8:21 AM EST

    More languages seem to be affected.

  4. identified · Feb 18, 8:24 AM EST

    We've rolled back to a recent deployment that affected the service. Service should be restored in a few minutes.

  5. monitoring · Feb 18, 8:28 AM EST

    Moving the incident to monitoring. We see that service is restored, requests and errors are back to pre-incident levels.

  6. resolved · Feb 18, 8:33 AM EST

    This incident has been resolved.

  7. postmortem · Feb 18, 8:34 AM EST

    A change to enable advanced debugging on a component supporting DeepL write for pro customers, caused the service to deploy incorrectly and reject incoming requests. We are looking into the root cause of this and will update this page with our findings after conducting a thorough internal review. We have identified the root cause and want to share the results here. The rollout of a feature to enhance error debugging resulted in a Write Pro instance being scheduled on an overloaded host, creating a "noisy neighbor" issue. This led to slow responses and connection pileups in our reverse proxy. To prevent similar incidents, we plan to implement several measures to better isolate resource-intensive workloads and improve how our services handle temporary slowdowns.

Updates are DeepL's own words from its status page (status: postmortem); we never add an inferred cause. The vendor's incident page ↗

Partial outage on DeepL Pro, DeepL Pro API

major
Started
Feb 18, 8:33 AM EST
Resolved
Feb 18, 9:10 AM EST
Duration
37 min

DeepL's updates (3)

  1. investigating · Feb 18, 8:33 AM EST

    We are investigating this issue.<br><br><b>Affected Services:</b> Document Translations<br><b>Additional Info:</b> We are investigating increased errors when translating documents

  2. resolved · Feb 18, 9:10 AM EST

    This incident has been resolved.

  3. postmortem · Feb 19, 5:30 AM EST

    We were experiencing performance issues when we migrated a message queuing node to our new network. This affected clients consuming from partitions on that node, particularly Document Translation services. The slowdown occurred because the message queue had to replicate partitions to the rejoined node-a resource-intensive process given our data volume. We've since implemented improvements, including scheduling migrations during off-peak hours and improving replication monitoring. We apologize for the disruption and are committed to preventing similar incidents.

Updates are DeepL's own words from its status page (status: postmortem); we never add an inferred cause. The vendor's incident page ↗

AIDownCheck is not affiliated with DeepL. This recap summarizes incidents published on DeepL's official status page.