Welcome to the Alteryx Knowledge Base

Knowledge Base / Server / Unable to upgrade Server with User managed database on version 4.2+

Unable to upgrade Server with User managed database on version 4.2+

user

Created/Edited - 12/6/2024 by Mitesh Narottam | Alteryx

Summary

Customer unable to upgrade user managed database on version 4.2+. 4.2+ Is currently not supported however the following steps may help to resolve this problem

Environment

The Alteryx service fails to start after upgrading Alteryx server on a user managed mongo database on version 4.2+. The last startup error show the following error/s:

Mongo error: "Index with name: AS_Applica_1_Type_1_UserName_1_CreationDa_1_Status_1_Owner_1_Completion_1_QOS_Priori_1_AppName_1 already exists with a different name: generic server error" code: <mongodb:85>

Mongo error: "Index with name: fd5e7f7ad987788251ca676578977ec1 already exists with a different name: generic server error" code: <mongodb:85>

Alteryx Server
- Version (Any)
MongoDB User Managed
- Version 4.2+

Cause

This occurs due to changes with indexing options in later version of MongoDB (See MongoDB Article ). The behaviour after upgrading is causing the index name field to concatenate attributes with field names instead of the index names provided by Alteryx. These values can been seen in Alteryx Service database and usually impact the the following collections. The expected index names are as follows:

AS_Applications = 6aa7e510a23f4c9639c519975d1f79c6
AS_Insights = 35d633d6cc212a8f63294ed8cdd62f78
AS_Queue = fd5e7f7ad987788251ca676578977ec1
AS_QueueInputs = 6faf2b7d081ecc6afe51272c41b23f59
AS_Results = 117c60e1e2b1cabd69047be9fc755fb6
AS_ResultsFiles = 93f54ef124c56e82b083cb1c5f28b8ba
AS_Schedules = cb8d7dd41a10a8ee018a0802f2efb978
AS_TileSetAccessTimes = cac478c3bdb039b6b8f19c37d91d82dc

Resolution

Please note: This process has not been officially tested and highly recommend taking a backup prior to making any changes (Click here for backup procedure). We do not officially support MongoDB v4.2.x at this time and this article is intended to assist the customer with resolving this particular issue. The highest MongoDB version currently supported is version 4.0. The customer should consider rolling back to last supported version avoid any future issue. That being said see resolution steps below:

Solution

If Alteryx has already been upgraded, Rollback to the last working version (see backup and recovery guide). This scenario rolls back to 2019.4.8
Stop the Alteryx Service on the Controller and any worker nodes
Attempt the Alteryx server upgrade again. (Please ensure the Alteryx Service does not start after the upgrade)
Connect to the Alteryx Service Database using Robo3t (How to connect to user-managed Mongo using Robo3t ).

Since its user managed, MongoDB is likely to be running independently from the Alteryx Service on another server. If for any reason you need to start MongoDB Service, click here for guidence
Please ensure the account used in Alteryx System Settings as the ability to recreate indexes.

Next we need to drop collection indexes. Right Click on Alteryx Service DB and select Open Shell. Then paste the following and click Run.

use AlteryxService
db.AS_Queue.dropIndexes()
db.AS_Results.dropIndexes()
db.AS_ResultsFiles.dropIndexes()
db.AS_QueueInputs.dropIndexes()
db.AS_Schedules.dropIndexes()
db.AS_Insights.dropIndexes()
db.AS_Applications.dropIndexes()

Check each collection tab and indexes have dropped.

Start the Alteryx Service on Controller and Worker nodes. Alteryx should rebuild the index name with the correct parameters.

Note: You can run the same commands directly from Mongo Shell. (How to connect to user managed DB via shell ). Simply connect to the Alteryx Service DB and Run the following

use AlteryxService
db.AS_Queue.dropIndexes()
db.AS_Results.dropIndexes()
db.AS_ResultsFiles.dropIndexes()
db.AS_QueueInputs.dropIndexes()
db.AS_Schedules.dropIndexes()
db.AS_Insights.dropIndexes()
db.AS_Applications.dropIndexes()

Additional Resources

Was this article helpful?