When creating a tenant in a ESB worker/manager separated setup, fronted by
an ELB, I see the same behavior (A dead page is returned once the tenant is
created and we see the INFO message - CommonUtil A tenant with same domain
already exist. Please use a different domain name. tenant domain: evan.com.
)

[1] - https://wso2.org/jira/browse/CARBON-14240

Regards,
Evanthika

On Wed, Jul 31, 2013 at 8:31 PM, Afkham Azeez <[email protected]> wrote:

>
>
>
> On Tue, Jul 30, 2013 at 9:24 AM, Nirodha Pramod <[email protected]> wrote:
>
>>
>>
>>
>> On Tue, Jul 30, 2013 at 9:20 AM, Nirmal Fernando <[email protected]> wrote:
>>
>>>
>>>
>>>
>>> On Tue, Jul 30, 2013 at 9:11 AM, Nirodha Pramod <[email protected]>wrote:
>>>
>>>> Hi Nirmal,
>>>>
>>>> On Tue, Jul 30, 2013 at 8:49 AM, Nirmal Fernando <[email protected]>wrote:
>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Tue, Jul 30, 2013 at 8:20 AM, Evanthika Amarasiri <
>>>>> [email protected]> wrote:
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> We noticed the error mentioned in $subject when creating tenants on
>>>>>> our App-server worker manager separated setup, fronted by an ELB. 
>>>>>> However,
>>>>>> the tenants do get created but they are not activated.
>>>>>>
>>>>>> We see the below exception at ELB console.
>>>>>>
>>>>>> TID: [0] [ELB] [2013-07-30 08:18:17,377]  WARN
>>>>>> {org.apache.synapse.transport.passthru.SourceHandler} -  Connection time
>>>>>> out after request is read: http-incoming-22
>>>>>> {org.apache.synapse.transport.passthru.SourceHandler}
>>>>>> TID: [0] [ELB] [2013-07-30 08:18:17,471]  WARN
>>>>>> {org.apache.synapse.transport.passthru.TargetHandler} -  
>>>>>> http-outgoing-20:
>>>>>> Connection time out while in state: REQUEST_DONE
>>>>>> {org.apache.synapse.transport.passthru.TargetHandler}
>>>>>> TID: [0] [ELB] [2013-07-30 08:18:18,472]  INFO
>>>>>> {org.wso2.carbon.lb.endpoint.endpoint.TenantAwareLoadBalanceEndpoint} -
>>>>>>  Failed over to Host:192.168.4.52, Remote Host:null, Port: 4250,
>>>>>> HTTP:10763, HTTPS:10443, Domain: as.qa.420.domain, Sub-domain:mgt,
>>>>>> Active:true
>>>>>> {org.wso2.carbon.lb.endpoint.endpoint.TenantAwareLoadBalanceEndpoint}
>>>>>>
>>>>>
>>>>> FYI this doesn't mean that the tenant is not activated, but for some
>>>>> reason connection to a member of the cluster got timed out and ELB failed
>>>>> over to another member of the cluster.
>>>>>
>>>>
>>>> Here the Tenant does not get activated, and we cannot login with the
>>>> created tenant. You have to activate the tenant manually to log in. A
>>>>
>>>
>>> Yes, this is a separate issue AFAIU.
>>>
>>>
>>>>  lso as we observed due to the time out it (probably ELB) tries to
>>>> send another tenant creation submit request which results in logging
>>>> "A tenant with same domain already exist" in AS side.
>>>>
>>>
>>>  Well, ELB doesn't send any such tenant creation requests AFAIK.
>>>
>> Well what I meant was a retry or something in that sort.(just guessed :) )
>>
>>
> Yes you are correct. On certain error codes, ELB fails over, and in this
> case, it is a timeout, and failover has happened. This timeout will go away
> once we fix the tenant creation time issue. We narrowed it down to the time
> it takes to copy the tenant's theme in its registry space. Senaka has
> committed a fix, which improves the tenant creation time.
>
> _______________________________________________
> Dev mailing list
> [email protected]
> http://wso2.org/cgi-bin/mailman/listinfo/dev
>
>
_______________________________________________
Dev mailing list
[email protected]
http://wso2.org/cgi-bin/mailman/listinfo/dev

Reply via email to