Alteryx Server Failover
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Notify Moderator
Is there any reason you cannot create a 2-node failover cluster for Alteryx Server?
In this example you would have the Controller, Gallery & Workers on both nodes (MongoDB being run externally on Mongo Atlas) with the MS Cluster Failover Roles configured to manage the Active Controller (following the HA instructions described here)
Solved! Go to Solution.
- Labels:
- Server
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Notify Moderator
Hi @paul_houghton ,
Alteryx failover follows general failover theory where you need to have a cluster quorum with an odd number of nodes. I'm not really an expert on the field and I don't read enough to explain why it is important to have a cluster quorum, but here are more information about it.
https://learn.microsoft.com/en-us/windows-server/failover-clustering/manage-cluster-quorum
About the other components, you don't have to follow the instructions, it is ok to have your MongoDB with replica sets (3 nodes as well 😁) and the gallery and worker components in different servers if you want to. Of course, for each of the nodes, you would need an additional server license.
Best,
Fernando Vizcaino
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Notify Moderator
That makes sense and confirms my thoughts.
The cluster quorum requirement can be satisfied with a witness file, which can be a dynamic and cloud-based witness. The Gallery and worker can be installed on both nodes with the controller then, should any one node go down the Server would still be available.
We would use Mongo Atlas for this application, so there is no need to manage the DB.