[ 
https://issues.apache.org/jira/browse/UNOMI-770?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Kevan Jahanshahi reassigned UNOMI-770:
--------------------------------------

    Assignee: Kevan Jahanshahi

> Migration of event rollover can be performed twice
> --------------------------------------------------
>
>                 Key: UNOMI-770
>                 URL: https://issues.apache.org/jira/browse/UNOMI-770
>             Project: Apache Unomi
>          Issue Type: Bug
>    Affects Versions: unomi-2.2.0
>            Reporter: Kevan Jahanshahi
>            Assignee: Kevan Jahanshahi
>            Priority: Major
>             Fix For: unomi-2.3.0
>
>
> The migration system is maintaining an history of executed steps to be able 
> to recover in case of error. But we have an issue for rollover indices the 
> migration step key is named:
>  
> {code:java}
> Migration step: Reindex step for: context-event-date-2017-05 (clone creation) 
> reach: STARTED 
> 2023-04-20T12:20:05,345 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  
> Migration step: Reindex step for: context-event-date-2017-05 (clone creation) 
> reach: COMPLETED 
> 2023-04-20T12:20:05,346 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  
> Migration step: Reindex step for: context-event-date-2017-05 (recreate the 
> index and perform the re-indexation) reach: STARTED 
> 2023-04-20T12:20:06,592 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  
> Migration step: Reindex step for: context-event-date-2017-05 (recreate the 
> index and perform the re-indexation) reach: COMPLETED 
> 2023-04-20T12:20:06,593 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  
> Migration step: Reindex step for: context-event-date-2017-05 (delete clone) 
> reach: STARTED 
> 2023-04-20T12:20:06,868 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  
> Migration step: Reindex step for: context-event-date-2017-05 (delete clone) 
> reach: COMPLETED 
> 2023-04-20T12:20:06,869 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  
> Migration step: Reindex step for: context-event-date-2017-05 (refresh at the 
> end) reach: STARTED 
> 2023-04-20T12:20:06,906 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  ES 
> Cluster status is yellow 
> 2023-04-20T12:20:06,907 | INFO  | features-3-thread-1 | MigrationContext      
>            | 238 - org.apache.unomi.shell-commands - 2.3.0.SNAPSHOT |  
> Migration step: Reindex step for: context-event-date-2017-05 (refresh at the 
> end) reach: COMPLETED  {code}
> So in case we restart the migration for any reasons the newly created events 
> indices will be migrated: 
> {code:java}
> Migration step: Reindex step for: context-event-000001 (clone creation) 
> reach: STARTED  {code}
> And this should not happen, since they already have been migrated.
> We need to improve the step keys to avoid newly indices being reindexed. (or 
> more easy we need to make sure the index name is not a rollover syntax.)



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to