Re: 11gR2 Standby Database : Throttle Archive logs

From: Nassyam Basha <nassyambasha_at_gmail.com>
Date: Wed, 12 Jun 2013 21:43:49 +0530
Message-ID: <CAABGLuJEr=MB6u-E-e+651sSUEZuOE-6Gfj2M7mmf36s1Axcjw_at_mail.gmail.com>

Well said, DELAY parameter can lag when applying. Then Bala should consider to tune the unit of data you sending/receving by configuring SDU size. And the socket buffer size can be configured in sqlnet.ora file and the value can be estimated by using the formula "Socket buffer size = 3 * (Bandwidth Speed) * (RTT)" . If you configure socket buffer size it will applicable to all incoming connections.
Apart from that you can also configure in each Oracle net service as below. Standby =
  (DESCRIPTION =

    (RECV_BUF_SIZE=11718750)
    (SEND_BUF_SIZE=11718750)
    (ADDRESS = (PROTOCOL = TCP)(HOST = 111.111.11.11)(PORT = 1521))
    (CONNECT_DATA =
      (SERVER = DEDICATED)
      (SERVICE_NAME = Standby)

   )
  )

For more details you can check the document http://www.oracle.com/au/products/database/maa-wp-10gr2-dataguardnetworkbestpr-134557.pdf

Also refer below link for "*Network Bandwidth Implications of Oracle Data Guard *"
http://www.oracle.com/technetwork/database/features/availability/dataguardnetwork-092224.html

But this may not work exactly in your enviornment, Because you would like to increase in offpeak hours, decrease shipment in peak hours.

Thanks & Regards,
Nassyam Basha.

On Wed, Jun 12, 2013 at 9:36 PM, Bala <oratips_at_gmail.com> wrote:

> Thank you Raj,Peter, Dimitre for pointing me to various options.. I will
> take it from here and get back to the list if i find anything interesting
>
> Thanks again!
> Bala
>
>
>
> On Wed, Jun 12, 2013 at 11:54 AM, Peter Gram Miracle A/S <pgr_at_miracleas.dk
> > wrote:
>
>> The DELAY parameter is only a delay in the applying of the redo data on
>> the
>> standby database and not in the transport process. have a look in chapter
>> 7.2.2 of the Oracle Data Guard Concepts and Admin
>> /Gram
>>
>>
>> Best Regards / Venlig Hilsen
>>
>> Peter Gram
>>
>> Miracle Consulting <http://www.miracleas.dk>
>>
>> Borupvang 2c, 2750 Ballerup, Denmark
>> Cell: (+45) 53747107 Home: (+45) 38745696
>> Office Phone: (+45) 44668855 Office Fax: (+45) 44668856
>> Mail: peter.gram_at_miracleas.dk
>> OakTable member <http://www.oaktable.net> <http://www.oaktable.com/>
>> ----------------------------------------------------
>>
>>
>>
>>
>>
>> On 12 June 2013 17:36, Nassyam Basha <nassyambasha_at_gmail.com> wrote:
>>
>> > Bala,
>> > You can also configure DELAY parameter for remote
>> > destination(log_archive_dest_n), So that you can delay the redo
>> transport
>> > to standby destinations and you can check the delay from
>> >
>> > SQL> SELECT DELAY_MINS FROM V$ARCHIVE_DEST;
>> >
>> > compression is again extra license with advanced compression.
>> >
>> > Thanks & Regards,
>> > Nassyam Basha
>> > On Jun 12, 2013 8:46 PM, "Bala" <oratips_at_gmail.com> wrote:
>> >
>> > > Gurus,
>> > > Can we throttle the shipment of Archive logs to occur at different
>> times
>> > > (increase in offpeak hours, decrease shipment in peak hrs) standing up
>> > > 11gR2 Datagaurd ?
>> > > The goal is to being sensitive to Network Bandwidth.
>> > >
>> > > Any insight will be greatly appreciated
>> > >
>> > > Thanks in advance.
>> > >
>> > > Best
>> > > Bala
>> > >
>> > > --
>> > > Bala Rao
>> > >
>> > >
>> > > --
>> > > http://www.freelists.org/webpage/oracle-l
>> > >
>> > >
>> > >
>> >
>> >
>> > --
>> > http://www.freelists.org/webpage/oracle-l
>> >
>> >
>> >
>>
>>
>> --
>> http://www.freelists.org/webpage/oracle-l
>>
>>
>>
>
>
> --
> Bala Rao
>

-- 
Thanks & Regards,
Nassyam Basha.


--
http://www.freelists.org/webpage/oracle-l

Received on Wed Jun 12 2013 - 18:13:49 CEST