Visualizations and Dashboard Configured in kibana 6.6.1 is not available in kibana 7.0.1 after upgrading ELK Stack from 6.6.1 to 7.0.1

[Varun_S]

Hi

1. I have deployed Elk stack of 6.6.1 and everything is fine like cluster health and there are no unassigned Shards. Also we have configured visualizations and Dashboards in kibana.

2. Now i am trying to upgrade Elk stack from 6.6.1 to 7.0.1. Upgrade happened Successfully. We are able to see all indices and Data and cluster_state is green, But visualizations and Dashboards configured in kibana 6.6.1 are not available in the kibana 7.0.1.

3. Can you please let us know what is the problem here.

[markov00 (Marco Vettorello)]

Hi @Varun_S
Thanks for your request. May I ask you what step you took to upgrade from 6.6.1. to 7.0.1?
The recommendation is to upgrade first to 6.7 that have a feature called Upgrade Assistant that helps you in the migration
https://www.elastic.co/guide/en/kibana/current/upgrade.html#_upgrading_from_6_x
usually visualization/dashboard/saved objects can be easily migrated from 6.x to 7.x.

Does the logs report anything about the errors in the migrations of visualization and objects?

One last question: are you using any third party plugins in Kibana that saves visualization and other saved objects?

[Varun_S]

HI @markov00,

Thanks for your reply.

1. I am deploying ELK Stack in Kubernets Environment via helm. So i used helm upgrade command to upgrade ELK Stack to 7.0.1.
2. I can't use Upgrade Assistant because i am using ELK Stack OSS
3. Once the upgrade is successful, I checked the kibana logs. Actually there should be migration of index from kibana_1 to kibana_2, But the logs says that kibana_1 index is getting created .Because of this Visualizations and Dashboards are getting Deleted. Please find the kibana logs for your reference.

Kibana logs:-

{"type":"log","@timestamp":"2019-07-31T11:38:44Z","tags":["info","migrations"],"pid":1,"message":"Creating index .kibana_1."}
{"type":"log","@timestamp":"2019-07-31T11:38:45Z","tags":["info","migrations"],"pid":1,"message":"Pointing alias .kibana to .kibana_1."}
{"type":"log","@timestamp":"2019-07-31T11:38:45Z","tags":["info","migrations"],"pid":1,"message":"Finished in 1439ms."}
{"type":"log","@timestamp":"2019-07-31T11:38:45Z","tags":["listening","info"],"pid":1,"message":"Server running at http://0:5601"}

Can you please help me here.

Regards,
Varun S

[markov00 (Marco Vettorello)]

Hey @michael.russell do you have any idea there?

[michael.russell]

I am deploying ELK Stack in Kubernets Environment via helm. So i used helm upgrade command to upgrade ELK Stack to 7.0.1.

Which helm chart are you using. Is it GitHub - elastic/helm-charts: You know, for Kubernetes or something else? In any case this doesn't appear to be a helm specific issue, more an issue with the upgrade procedure that you have taken.

Now i am trying to upgrade Elk stack from 6.6.1 to 7.0.1.

This is not a supported upgrade path. As per the upgrade docs you need to upgrade to 6.7 first. Some other really important steps to take are to test this first in a development environment and make a full cluster snapshot first so that you can restore if the upgrade doesn't work (Upgrade Elasticsearch | Elasticsearch Guide [7.0] | Elastic).

I think the next thing you can try is Troubleshooting saved object migrations | Kibana Guide [7.0] | Elastic. If this doesn't work the only option left is to restore from the latest snapshot.

[Varun_S]

Hi @michael.russell

Thanks for your reply i will check once from my side.

Regards,
Varun S