Qlik Sense works normally, but a huge quantity Repository logs is generated. These following messages are continuously repeated (not necessarily both...
Qlik Sense works normally, but a huge quantity Repository logs is generated. These following messages are continuously repeated (not necessarily both of them are present).
In System logs:
INFO SERVERNODE System.Repository.Repository.Core.Repository.Common.ModelRepository`1[[Repository.Domain.Model.UserDirectorySettings, Repository.Domain, Version=xx.xx.x.0, Culture=neutral, PublicKeyToken=...]]
..... Database object has newer timestamp than object being modified; ....
.... ERROR SERVERNODE System.Repository.Repository.Core.Repository.Common.TransactionUtility
..... Error when committing
In User Management logs:
WARN SERVERNODE UserManagement.Repository.Repository.Users.SyncUsersWorker ........
Conflict when trying to do database operation; retrying... at Repository.Core.Repository.Common.AbstractModelRepository.CheckConflicts(IRepositoryObject persistedItem, DateTime itemModifiedDate)↵
The quantity of messages generated is very high and several GB of memory can be occupied in disk in few hours.
- Qlik Sense until February 2021 Patch 6.
In order to avoid a complete filling of the disk, it is necessary to restart the node where the problem occurs. Generally, this resolves the issue.
To avoid this problem, we have introduced an improvement in the latest Qlik Sense releases that avoids potential conflicts. This improvement is present in Qlik Sense February 2021 Patch 6 and in Qlik Sense May 2021 SR1 or higher releases. An upgrade is recommended to ultimately fix the problem.
The problem is due to an incorrect synchronization between the Database and one of the nodes.
Internal Investigation ID(s)