Around 20:55 UTC on the 29th of September, Sigma users were unable to find via search workbooks created in the last 3 months, due to an inadvertent infrastructure change.
All customers on Azure, AWS (us-west2 and eu-central-1), and GCP (us-central-1), using the file explorer in embeds, using the Sigma API for workbooks and files, or searching across Sigma were affected.
We restored the search database to properly include the latest information in the search results.
Incident Start Time: Approximately 20:44 UTC September 29, 2023
Incident End Time: Approximately 23:17 UTC September 29, 2023
|2023-09-29 20:55:00||Incident alerted in our internal monitoring systems|
|2023-09-29 21:10:00||Incident noticed and flagged internally.|
|2023-09-29 21:38:00||Azure eastus2 resolved on its own due to a small number of search updates|
|2023-09-29 21:46:00||Incident escalated internally|
|2023-09-29 21:47:00||First customer report|
|2023-09-29 22:30:00||Customer issue linked to incident|
|2023-09-29 22:31:00||Incident escalated to highest priority|
|2023-09-29 22:42:00||Manually restored to prior search database state for GCP|
|2023-09-29 22:50:00||GCP us-central1 resolved|
|2023-09-29 23:00:00||Manually restored to prior search database state for AWS us-west-2|
|2023-09-29 23:07:00||AWS us-west-2 resolved|
|2023-09-29 23:11:00||Manually restored to prior search database state for AWS eu-central-1|
|2023-09-29 23:17:00||AWS eu-central-1 resolved|
|2023-09-29 23:17:00||Issue resolved|
A routine infrastructure change accidentally swapped the latest search database to a state from 3 months prior. This caused any changes in the past few months to not be visible in search.
This meant that any applications built on embedded Sigma which used the workbooks API for navigation failed to operate properly if they depended on documents newer than 3 months old.
All customers on Azure, AWS (us-west2 and eu-central-1) and GCP (us-central-1) using the file explorer in embeds, hitting the Sigma API for workbooks and files, or searching across Sigma were affected.
The search results returned to the correct state after rolling back to the previous healthy version of the search database.
- We’re looking to refine our escalation procedures in such cases for a faster resolution.
- Working to continuously improve our internal alerting system.
- Looking to put in place better guardrails to avoid such states.
- Replaced our mechanism for updating the search database state with a new mechanism which does not run the same risk of being triggered accidentally.
We apologize for the disruption and inconvenience that you have endured as a result of this situation. Your trust in us is of the utmost importance, and we are dedicated to taking the necessary measures to prevent similar incidents from occurring in the future. If you have any questions please reach out to our Support Team.
Thank you for your continuous support, patience and understanding.