Re: 2 standbys using dependency=

  • From: Howard Latham <howard.latham@xxxxxxxxx>
  • To: Subodh Deshpande <deshpande.subodh@xxxxxxxxx>
  • Date: Fri, 14 Jan 2011 15:57:44 +0000

This is the Doco


   -

   DEPENDENCY=LOG_ARCHIVE_DEST_n

   Defines one archival destination to receive redo data on behalf of
   several destinations. Redo transport services transmit redo data to the
   destination that then shares its archived redo log files with multiple
   standby databases.




On 14 January 2011 15:52, Howard Latham <howard.latham@xxxxxxxxx> wrote:

> BY using dependency= in log_archive_dest you get only one log on the remote
> box which both standbys use. This works very well but we get the temporary
> hang for log switch. Judging by the silence maybe no-one is using it.
>
> LAD_1=location=blah
> LAD_2=service=orcl async valid_for=(online_logfile,primary_role)
> db_unique_name=orcl reopen=30
> LAD_3=service=test6stby async
> valid_for=(online_logfile,primary_role)db_unique_name=test6stby
> dependency=log_archive_dest_2
>
>
>
>
>
>
> On 14 January 2011 15:42, Subodh Deshpande <deshpande.subodh@xxxxxxxxx>wrote:
>
>> hi
>> Quote
>> I want to have two standbys on the standby box but I don't want to ship
>> two copies of archived logs
>> UnQuote
>>
>> I am not sure whether this is possible, is ther any document that states
>> it is possible.
>>
>> casue according to me when you ship to standby and further when it gets
>> applied onto the standby database, relevant views get updated and the
>> purpose of received archive is over.
>>
>> I am not sure and you can try one workaround, which is, letus copy the
>> non-applied(just received copy) archive to some else destination on the same
>> host for your other standby.
>>
>> may you want to do this to reduce the network traffice and as well as one
>> more standby on the same host.
>>
>> it looks like usually RAC has one standby you are preparing RAC standby
>> for single node database..
>>
>> have I understood you proprely..!
>>
>> thanks..
>> subodh
>>
>>
>>
>>
>> On 14 January 2011 18:40, Howard Latham <howard.latham@xxxxxxxxx> wrote:
>>
>>> Oracle 11G RH LINUX
>>>
>>> I have logged an SR but so far nothing!( 20 hours)
>>>
>>> I have a primary and standby db Not using Dataguard Broker.
>>>
>>> I want to have two standbys on the standby box but I don't want to ship
>>> two copies of archived logs
>>>
>>> So I tried using  log_archive_dest_3=dependency=log_achive_dest_2
>>>
>>> The Primary stops archiving and stalls when the 3rd and last online redo
>>> log is full.
>>>
>>> The Dread message All logs need archiving appear in the alert log.
>>>
>>> The Database seems to start writing all three archived logs and hangs
>>> while it does it.
>>> If I shutdown the Standbys the Database appears to  free up.
>>>
>>> This is a Production DB so I cant play around very much with Parms and am
>>> thinking of manually maintaining the second Standby.
>>>
>>> Any suggestions Folks?
>>>
>>> LGR Processes =10
>>>
>>>
>>>
>>> --
>>> Howard A. Latham
>>>
>>> Sent from my Nokia N97
>>>
>>>
>>
>>
>> --
>> ==============================
>> DO NOT FORGET TO SMILE TODAY
>> ==============================
>>
>
>
>
> --
> Howard A. Latham
>
> Sent from my Nokia N97
>
>


-- 
Howard A. Latham

Sent from my Nokia N97

Other related posts: