Hi Raj,

I guess since the server start logic was moved to the CA, a start servers
extension will have to be implemented again. I will look in to this and
submit a fix.


Regards,
Chamila de Alwis
Software Engineer | WSO2 | +94772207163
Blog: code.chamiladealwis.com



On Mon, Mar 2, 2015 at 4:27 PM, Rajkumar Rajaratnam <[email protected]>
wrote:

> I have built php docker image locally and found that members are
> terminated on app undeployment. So the issue is resolved with the fix.
>
> However, members was not activated because port 80 was not open in
> containers (agent log says Port '80' is not active). Will look into it.
>
> Also wee need to build all cartridge-images with latest agent and push
> these to docker hub, otherwise app undeployment won't work.
>
> Thanks.
>
> On Mon, Mar 2, 2015 at 3:10 PM, Rajkumar Rajaratnam <[email protected]>
> wrote:
>
>> Thanks Chamila for the quick fix. I have now merged into master branch.
>>
>> Thanks.
>>
>> On Mon, Mar 2, 2015 at 2:58 PM, Chamila De Alwis <[email protected]>
>> wrote:
>>
>>> Hi Raj,
>>>
>>> This has been caused by an incorrect addition of the instanceClusterId
>>> from the received JSON object. I've submitted the fix at PR[1]. Please
>>> review and merge.
>>>
>>> [1] - https://github.com/apache/stratos/pull/248
>>>
>>>
>>> Regards,
>>> Chamila de Alwis
>>> Software Engineer | WSO2 | +94772207163
>>> Blog: code.chamiladealwis.com
>>>
>>>
>>>
>>> On Mon, Mar 2, 2015 at 2:45 PM, Chamila De Alwis <[email protected]>
>>> wrote:
>>>
>>>> Hi Raj,
>>>>
>>>> Yes, this is wrong. I'll correct this and submit a PR soon.
>>>>
>>>>
>>>> Regards,
>>>> Chamila de Alwis
>>>> Software Engineer | WSO2 | +94772207163
>>>> Blog: code.chamiladealwis.com
>>>>
>>>>
>>>>
>>>> On Mon, Mar 2, 2015 at 1:47 PM, Rajkumar Rajaratnam <[email protected]
>>>> > wrote:
>>>>
>>>>> Hi Devs,
>>>>>
>>>>> I am experiencing $Subject in kubernetes setup with
>>>>> stratos/php:4.1.0-beta docker image (probably outdated agent).
>>>>>
>>>>> Reason for this issue is that member termination cycle is broken at
>>>>> python agent.
>>>>>
>>>>> Agent is failing to process Instance Cleanup Cluster Event, so agent
>>>>> won't send instance ready to shutdown event CC, CC won't send member ready
>>>>> shutdown event to AS, AS won't ask CC to terminate the instance. So
>>>>> monitors and members are running even after we undeploy the application.
>>>>>
>>>>> Here is the agent log.
>>>>>
>>>>> [2015-03-02 07:51:29,971] DEBUG
>>>>> {agent.py:on_application_signup_removed} - Application signup removed 
>>>>> event
>>>>> received: '{"applicationId":"single-cartridge-app","tenantId":-1234}'
>>>>> [2015-03-02 07:51:29,971] ERROR
>>>>> {agent.py:on_application_signup_removed} - Error processing tenant
>>>>> unSubscribed event
>>>>> Traceback (most recent call last):
>>>>>   File "agent.py", line 347, in on_application_signup_removed
>>>>>
>>>>> CartridgeAgent.extension_handler.on_application_signup_removal_event(event_obj)
>>>>>   File
>>>>> "/mnt/apache-stratos-python-cartridge-agent-4.1.0-SNAPSHOT/modules/extensions/extensionhandler.py",
>>>>> line 362, in on_application_signup_removal_event
>>>>>     " [application ID] " +
>>>>> application_signup_removal_event.applicationId
>>>>> TypeError: cannot concatenate 'str' and 'int' objects
>>>>> [2015-03-02 07:51:30,002] DEBUG {eventsubscriber.py:on_message} -
>>>>> Message received: topology/ClusterInstanceTerminatingEvent:
>>>>>
>>>>> {"serviceName":"php","clusterId":"my-php.php.domain","appId":"single-cartridge-app","instanceId":"single-cartridge-app-1"}
>>>>> [2015-03-02 07:51:30,002] DEBUG {eventsubscriber.py:run} - Event
>>>>> handler not found for event : 'ClusterInstanceTerminatingEvent'
>>>>> [2015-03-02 07:51:30,013] DEBUG {eventsubscriber.py:on_message} -
>>>>> Message received: instance/notifier/InstanceCleanupClusterEvent:
>>>>>
>>>>> {"clusterId":"my-php.php.domain","clusterInstanceId":"single-cartridge-app-1"}
>>>>> [2015-03-02 07:51:30,014] DEBUG {eventsubscriber.py:run} - Executing
>>>>> handler for event 'InstanceCleanupClusterEvent'
>>>>> [2015-03-02 07:51:30,014] ERROR {eventsubscriber.py:run} - Error
>>>>> processing 'InstanceCleanupClusterEvent' event
>>>>> Traceback (most recent call last):
>>>>>   File
>>>>> "/mnt/apache-stratos-python-cartridge-agent-4.1.0-SNAPSHOT/modules/subscriber/eventsubscriber.py",
>>>>> line 103, in run
>>>>>     handler(event_msg)
>>>>>   File "agent.py", line 195, in on_instance_cleanup_cluster
>>>>>     event_obj =
>>>>> InstanceCleanupClusterEvent.create_from_json(msg.payload)
>>>>>   File
>>>>> "/mnt/apache-stratos-python-cartridge-agent-4.1.0-SNAPSHOT/modules/event/instance/notifier/events.py",
>>>>> line 67, in create_from_json
>>>>>     instance_cluster_id = json_obj["instanceClusterId"] if "clusterId"
>>>>> in json_obj else None
>>>>> KeyError: 'instanceClusterId'
>>>>>
>>>>>
>>>>> @Chamila, do you know the reason for this?
>>>>>
>>>>> --
>>>>> Rajkumar Rajaratnam
>>>>> Committer & PMC Member, Apache Stratos
>>>>> Software Engineer, WSO2
>>>>>
>>>>> Mobile : +94777568639
>>>>> Blog : rajkumarr.com
>>>>>
>>>>
>>>>
>>>
>>
>>
>> --
>> Rajkumar Rajaratnam
>> Committer & PMC Member, Apache Stratos
>> Software Engineer, WSO2
>>
>> Mobile : +94777568639
>> Blog : rajkumarr.com
>>
>
>
>
> --
> Rajkumar Rajaratnam
> Committer & PMC Member, Apache Stratos
> Software Engineer, WSO2
>
> Mobile : +94777568639
> Blog : rajkumarr.com
>

Reply via email to