11 KiB
| title | slug |
|---|---|
| Upgrade on Kubernetes | /deployment/upgrade/kubernetes |
Upgrade on Kubernetes
This guide will help you upgrade your OpenMetadata Kubernetes Application with automated helm hooks.
Requirements
This guide assumes that you have an OpenMetadata deployment that you installed and configured following the Kubernetes Deployment guide.
We also assume that your helm chart release names are openmetadata and openmetadata-dependencies and namespace used is
default.
Procedure
Below document is valid for upgrading Helm Charts from 0.11.5 to 0.12.X.
Back up metadata
Before proceeding, pleae make sure you made a backup of your MySQL/Postgres DB behind OpenMetadata server. This step is extremely important for you to restore to your current state if any issues come up during the upgrade
Learn how to back up MySQL data.Get an overview of what has changed in Helm Values
You can get changes from artifact hub of openmetadata helm chart release. Click on Default Values >> Compare to Version.
Upgrade Helm Repository with a new release
Update Helm Chart Locally for OpenMetadata with the below command:
helm repo update open-metadata
It will result in the below output on screen.
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "open-metadata" chart repository
Update Complete. ⎈Happy Helming!⎈
Verify with the below command to see the latest release available locally.
helm search repo open-metadata --versions
> NAME CHART VERSION APP VERSION DESCRIPTION
open-metadata/openmetadata 0.0.39 0.12.1 A Helm chart for OpenMetadata on Kubernetes
open-metadata/openmetadata 0.0.38 0.12.0 A Helm chart for OpenMetadata on Kubernetes
open-metadata/openmetadata 0.0.37 0.12.0 A Helm chart for OpenMetadata on Kubernetes
open-metadata/openmetadata 0.0.36 0.12.0 A Helm chart for OpenMetadata on Kubernetes
open-metadata/openmetadata 0.0.35 0.11.5 A Helm chart for OpenMetadata on Kubernetes
open-metadata/openmetadata 0.0.34 0.11.4 A Helm chart for OpenMetadata on Kubernetes
...
open-metadata/openmetadata-dependencies 0.0.39 0.12.1 Helm Dependencies for OpenMetadata
open-metadata/openmetadata-dependencies 0.0.38 0.12.0 Helm Dependencies for OpenMetadata
open-metadata/openmetadata-dependencies 0.0.37 0.12.0 Helm Dependencies for OpenMetadata
open-metadata/openmetadata-dependencies 0.0.36 0.12.0 Helm Dependencies for OpenMetadata
open-metadata/openmetadata-dependencies 0.0.35 0.11.5 Helm Dependencies for OpenMetadata
open-metadata/openmetadata-dependencies 0.0.34 0.11.4 Helm Dependencies for OpenMetadata
...
Upgrade OpenMetadata Dependencies
Step 1: Upgrade Airfow
We have upgraded the Airflow version from 2.1.4 to 2.3.3 with OpenMetadata 0.12.X releases.
Before you start upgrading OpenMetadata Dependencies, you must follow airflow migration docs here to upgrade Airflow.
Step 2: Upgrade OpenMetadata Dependencies with the below command
helm upgrade openmetadata-dependencies open-metadata/openmetadata-dependencies
The above command uses configurations defined here.
You can modify any configuration and deploy by passing your own values.yaml.
Step 3: Troubleshooting
If your helm upgrade fails with the below command result -
Error: UPGRADE FAILED: cannot patch "mysql" with kind StatefulSet: StatefulSet.apps "mysql" is invalid: spec: Forbidden: updates to statefulset spec for fields other than 'replicas', 'template', 'updateStrategy', 'persistentVolumeClaimRetentionPolicy' and 'minReadySeconds' are forbidden
This is probably because with 0.12.1, we have default size of mysql persistence set to 50Gi.
Kubernetes does not allow changes to Persistent volume with helm upgrades.
In order to work around this issue, you can either default the persistence size to 8Gi or run the below command which will patch Persistent Volumes and Persistent Volume Claims for mysql helm and then run the above helm upgrade command.
kubectl patch pvc data-mysql-0 -p '{"spec":{"resources":{"requests":{"storage":"50Gi"}}}}'
kubectl patch pv <mysql-pv> -p '{"spec":{"storage":"50Gi"}}'
Make sure that, when using your own values.yaml, you are not overwriting elements such as the image of the containers.
This would prevent your new deployment to use the latest containers when running the upgrade.
If you are running into any issues, double-check what are the default values of the helm revision.
Upgrade OpenMetdata
We upgrade OpenMetadata with the below command:
helm upgrade openmetadata open-metadata/openmetadata
You might need to pass your own values.yaml with the --values flag
Reindex ElasticSearch
We have added a conditional suggestion mapping for all of the elasticsearch indexes. This may require re-indexing. With 0.12.1 its never been easier to index your metadata
Go to Settings -> Event Publishers -> ElasticSearch
Make sure you select "Recreate Indexes"
Click on the "Recreate Indexes" lable and click "Re Index All"
Troubleshooting for 0.12 Release
Using custom helm values
If you are facing an issue similar to below when openmetadata pod keeps on restarting.
java.lang.ClassNotFoundException: org.openmetadata.catalog.security.DefaultAuthorizer
at java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:581)
at java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:178)
at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:522)
at java.base/java.lang.Class.forName0(Native Method)
at java.base/java.lang.Class.forName(Class.java:315)
at org.openmetadata.service.OpenMetadataApplication.registerAuthorizer(OpenMetadataApplication.java:240)
at org.openmetadata.service.OpenMetadataApplication.run(OpenMetadataApplication.java:123)
at org.openmetadata.service.OpenMetadataApplication.run(OpenMetadataApplication.java:92)
at io.dropwizard.cli.EnvironmentCommand.run(EnvironmentCommand.java:59)
at io.dropwizard.cli.ConfiguredCommand.run(ConfiguredCommand.java:98)
at io.dropwizard.cli.Cli.run(Cli.java:78)
at io.dropwizard.Application.run(Application.java:94)
at org.openmetadata.service.OpenMetadataApplication.main(OpenMetadataApplication.java:323)
The root cause of the issue is the default helm values which are upgraded in helm charts but are getting overridden by your custom helm values. Please verify the config for Authorizer Class Name and Container Request Filter.
We have changed org.openmetadata.catalog.security.* to org.openmetadata.service.security.*.
Make sure to verify your helm values and update the below content.
global:
...
authorizer:
className: "org.openmetadata.service.security.DefaultAuthorizer"
containerRequestFilter: "org.openmetadata.service.security.JwtFilter"
...
Troubleshooting for 0.10 Release
If you are upgrading from previous releases to 0.10, you might encounter the below logs which indicates that the 0.10 release is backward incompatible.
[I/O dispatcher 1] DEBUG org.apache.http.impl.nio.client.InternalIODispatch - http-outgoing-0 [ACTIVE] [content length: 263; pos: 263; completed: true]
[main] DEBUG org.elasticsearch.client.RestClient - request [PUT http://elasticsearch:9200/glossary_search_index/_mapping?master_timeout=30s&ignore_unavailable=false&expand_wildcards=open%2Cclosed&allow_no_indices=false&ignore_throttled=false&timeout=30s] returned [HTTP/1.1 400 Bad Request]
[main] ERROR org.openmetadata.catalog.elasticsearch.ElasticSearchIndexDefinition - Failed to update Elastic Search indexes due to
org.elasticsearch.ElasticsearchStatusException: Elasticsearch exception [type=illegal_argument_exception, reason=can't merge a non object mapping [owner] with an object mapping]
at org.elasticsearch.rest.BytesRestResponse.errorFromXContent(BytesRestResponse.java:176)
at org.elasticsearch.client.RestHighLevelClient.parseEntity(RestHighLevelClient.java:1933)
at org.elasticsearch.client.RestHighLevelClient.parseResponseException(RestHighLevelClient.java:1910)
at org.elasticsearch.client.RestHighLevelClient.internalPerformRequest(RestHighLevelClient.java:1667)
at org.elasticsearch.client.RestHighLevelClient.performRequest(RestHighLevelClient.java:1639)
at org.elasticsearch.client.RestHighLevelClient.performRequestAndParseEntity(RestHighLevelClient.java:1606)
at org.elasticsearch.client.IndicesClient.putMapping(IndicesClient.java:342)
at org.openmetadata.catalog.elasticsearch.ElasticSearchIndexDefinition.updateIndex(ElasticSearchIndexDefinition.java:139)
at org.openmetadata.catalog.elasticsearch.ElasticSearchIndexDefinition.updateIndexes(ElasticSearchIndexDefinition.java:91)
at org.openmetadata.catalog.util.TablesInitializer.execute(TablesInitializer.java:227)
at org.openmetadata.catalog.util.TablesInitializer.main(TablesInitializer.java:149)
Suppressed: org.elasticsearch.client.ResponseException: method [PUT], host [http://elasticsearch:9200], URI [/glossary_search_index/_mapping?master_timeout=30s&ignore_unavailable=false&expand_wildcards=open%2Cclosed&allow_no_indices=false&ignore_throttled=false&timeout=30s], status line [HTTP/1.1 400 Bad Request]
{"error":{"root_cause":[{"type":"illegal_argument_exception","reason":"can't merge a non object mapping [owner] with an object mapping"}],"type":"illegal_argument_exception","reason":"can't merge a non object mapping [owner] with an object mapping"},"status":400}
at org.elasticsearch.client.RestClient.convertResponse(RestClient.java:326)
at org.elasticsearch.client.RestClient.performRequest(RestClient.java:296)
at org.elasticsearch.client.RestClient.performRequest(RestClient.java:270)
at org.elasticsearch.client.RestHighLevelClient.internalPerformRequest(RestHighLevelClient.java:1654)
... 7 common frames omitted
Note that to migrate, you'll need a fresh installation! Follow the upgrade guide for more details.