[ 
https://issues.apache.org/jira/browse/GEODE-4893?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Jinmei Liao updated GEODE-4893:
-------------------------------
    Description: 
`import cluster-configuration --zip-file-name=...` in a cluster which has 
locator and servers (Server has no application regions defined yet). the server 
is shutdown without any message in the logs as to why the server was shutdown. 
Gfsh says import was successful.

On manual restart of the server I can see that the config was actually imported 
successfully.

 

We had added the feature that user can do an {{import cluster-configuration 
--zip-file-name=...}} without shutting down the vanilla servers, and the 
servers cache would just bounce. We should
 # make sure the server's cache is completely empty. i.e. not just without 
regions, should not have any other entities like async event queue, disk stores 
etc.
 # make sure after the import, the servers are not shut down.

We have ClusterConfigImportDUnitTest.testImportWithRunningServer to test this 
scenaio, in the Dunit environment, the servers are not shut down after the 
import. We will need to write a similar acceptance tests to expose the problem.

  was:
`import cluster-configuration --zip-file-name=...` in a cluster which has 
locator and servers (Server has no application regions defined yet). the server 
is shutdown without any message in the logs as to why the server was shutdown. 
Gfsh says import was successful.

On manual restart of the server I can see that the config was actually imported 
successfully.

 

We had added the feature that user can do an {{import cluster-configuration 
--zip-file-name=...}} without shutting down the vanilla servers, and the 
servers cache would just bounce. We should
 # make sure the server's cache is completely empty. i.e. not just without 
regions, should not have any other entities like async event queue, disk stores 
etc.
 # make sure after the import, the servers are not shut down.


> import cluster configuration command shuts down the existing servers
> --------------------------------------------------------------------
>
>                 Key: GEODE-4893
>                 URL: https://issues.apache.org/jira/browse/GEODE-4893
>             Project: Geode
>          Issue Type: Bug
>          Components: configuration, docs, gfsh
>    Affects Versions: 1.4.0
>            Reporter: Jinmei Liao
>            Priority: Major
>
> `import cluster-configuration --zip-file-name=...` in a cluster which has 
> locator and servers (Server has no application regions defined yet). the 
> server is shutdown without any message in the logs as to why the server was 
> shutdown. Gfsh says import was successful.
> On manual restart of the server I can see that the config was actually 
> imported successfully.
>  
> We had added the feature that user can do an {{import cluster-configuration 
> --zip-file-name=...}} without shutting down the vanilla servers, and the 
> servers cache would just bounce. We should
>  # make sure the server's cache is completely empty. i.e. not just without 
> regions, should not have any other entities like async event queue, disk 
> stores etc.
>  # make sure after the import, the servers are not shut down.
> We have ClusterConfigImportDUnitTest.testImportWithRunningServer to test this 
> scenaio, in the Dunit environment, the servers are not shut down after the 
> import. We will need to write a similar acceptance tests to expose the 
> problem.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to