Graylog & Opensearch error

I am attempting to migrate from Elasticserach 6.8 to Opensearch 1.3. When starting Graylog I get this error:


[2023-01-12T11:45:59,391][ERROR][o.o.b.Bootstrap          ] [graylog-es1] Exception
java.lang.IllegalStateException: The index [[.tasks/5gaAnPefT6S9oBzrfGKQsw]] was created with version [5.6.16] but the minimum compatible version is [6.0.0-beta1]. It should be re-indexed in OpenSearch 6.x before upgrading to 1.3.7.

Using this curl command I found this line in the output which seems to show which index is the issue.

curl -X GET "<ip>:9200/*/_settings"

".tasks":{"settings":{"index":{"creation_date_string":"2021-02-10T10:13:06.078Z","number_of_shards":"1","auto_expand_replicas":"0-1","provided_name":".tasks","creation_date":"1612951986078","priority":"2147483647","number_of_replicas":"1","uuid":"5gaAnPefT6S9oBzrfGKQsw","version":{"upgraded":"6082399","created_string":"5.6.16","upgraded_string":"6.8.23","created":"5061699"}}}}

CentOS 7
Graylog 4.3.9

Unfortunately I know little about Elasticsearch / Opensearch, so I’m not sure what the options are. My google skills have let me down so far.

Thanks

Hey, Richard, I wanted to reach out to you with some observations from the documentation. My response is based solely on the documentation. We have several other Graylog members who can response here with experience, so I’m sure they’re contribute, too.

Your error message indicates that the index ‘.tasks/5gaAnPefT6S9oBzrfGKQsw’ was created with version 5.6.16 of Elasticsearch, but the minimum compatible version for your current version of Graylog (4.3.9) is 6.0.0-beta1. So, to resolve this issue, you will need to re-index the ‘.tasks’ index in OpenSearch 6.x before upgrading to version 1.3.7 of OpenSearch.

This can be done by using the re-index API of Elasticsearch. You can also use a tool such as Elasticsearch-dump to backup and restore the index.

Also, please note that migrating from Elasticsearch 6.8 to OpenSearch 1.3 can be a major upgrade, so please consult the official documentation and seek help from our experienced Elasticsearch/OpenSearch users in the community. Good luck and thanks for asking!

3 Likes