Rolling Upgrade Scenario: Two OneSpan Authentication Server Instances Using Individual Databases
- 08 Jan 2025
- 1 Minute à lire
- SombreLumière
- PDF
Rolling Upgrade Scenario: Two OneSpan Authentication Server Instances Using Individual Databases
- Mis à jour le 08 Jan 2025
- 1 Minute à lire
- SombreLumière
- PDF
The content is currently unavailable in French. You are viewing the default English version.
Résumé de l’article
Avez-vous trouvé ce résumé utile ?
Merci pour vos commentaires
About this scenario
This scenario describes an environment with the following setup:
- There are two OneSpan Authentication Server instances: server A and server B.
- Each OneSpan Authentication Server instance has its own database.
- User load is distributed evenly between all OneSpan Authentication Server instances using a third-party solution.
- Two-way replication is enabled between all instances to ensure consistency of data across databases
Figure: Rolling upgrade scenario: Two servers using individual databases and replication
Before you begin
Verify that you have addressed the different usability and user load issues related to a rolling upgrade (see General considerations).
Walkthrough: Performing a rolling upgrade on two servers with individual databases
Performing a rolling upgrade on two servers with individual databases
- Remove the load from server A.
- Break replication from server B to server A.
- Wait until the replication queue on server A is empty.
- Stop the OneSpan Authentication Server service on server A.
- Break replication from server A to server B.
- Upgrade OneSpan Authentication Server on server A.
- Restore replication from server B to server A.
- Wait until the replication queue on server B is empty.
- Restore the load to server A.
- Remove the load from server B.
- Wait until the replication queue on server B is empty.
- Stop the OneSpan Authentication Server service on server B.
- Break replication from server B to server A.
- Upgrade OneSpan Authentication Server on server B.
- Restore replication from server A to server B.
- Restore replication from server B to server A.
- Wait until the replication queue on server A is empty.
- Restore load to server B.
Cet article vous a-t-il été utile ?