Re: Golden Gate Replicat behaving differently in different environment

From: Ravi Teja Bellamkonda <raviteja.bellamkonda7_at_gmail.com>
Date: Sun, 17 Dec 2017 14:14:23 -0800
Message-ID: <CANRhLpAPdWVJNbv8OFcUtKP8MaaspuxoeS4XFXFEz=hJPqiuNg_at_mail.gmail.com>



Thanks a lot. Will try to troubleshoot from OS.

On Dec 17, 2017 10:23 AM, "Bobby Curtis" <curtisbl_at_gmail.com> wrote:

> The AWR will just show you what SQL is being ran against the database. If
> you were using integrated products (extract/replication), then the AWR may
> be of some help; in this case not so much.
>
> I would start looking at your checkpoints for the extract and replicat
> (info extract <name>, show how). Then make sure data is being written to
> the trail file and that the replicat is not at the end of a file.
>
> Using classic processes (extract/replicat), you need to troubleshoot from
> the OS side. These processes rely heavily on the OS memory as well.
>
> Bobby Curtis
> Sent from my iPad Pro
>
> On Dec 17, 2017, at 10:38 AM, Ravi Teja Bellamkonda <
> raviteja.bellamkonda7_at_gmail.com> wrote:
>
> Is this a hub-n-spoke architecture?
>
> Yes, this is a hub-n-spoke architecture.
>
> Database A -> GG Server -> Database B.
>
>
> Are the processes integrated?
>
> No these are not integrated processes.
>
> Does this system do the same thing as the system you are comparing to?
>
> Yes, Both the systems are supporting same applications.
>
> Are there any data types that could be causing issues?
>
> We have lot of DML Operations on a table with CLOBs. But it is still the
> same on both environments.
>
> Snippet of Target AWR for environment in which replicat is slow,
>
> <image.png>
>
> The first 3 SQL are Insert, Delete and Update respectively.
>
>
> The following is the snippet where Replicat is processing faster
> comparatively.
>
> <image.png>
>
> This is odd that no of executions are high in this target which is a bit
> confusing for me. Please let me know any more info is needed.
>
> Thanks a lot for your time.
>
> On Dec 17, 2017 7:23 AM, "Bobby Curtis" <curtisbl_at_gmail.com> wrote:
>
>> Is this a hub-n-spoke architecture?
>> Are the processes integrated?
>> Does this system do the same thing as the system you are comparing to?
>> Are there any data types that could be causing issues?
>>
>> These are just some of the basic questions. A bit more information is
>> needed before a comparison can be made.
>>
>> Thanks
>> Bobby Curtis
>> http://about.me/dbasolved
>> Sent from my iPhone
>>
>> > On Dec 17, 2017, at 09:52, Ravi Teja Bellamkonda <
>> raviteja.bellamkonda7_at_gmail.com> wrote:
>> >
>> > Hi List,
>> >
>> > We have a GG Setup where extract and replicat processes are running
>> remotely.
>> >
>> > GG Version: 11.2.1.0.3
>> >
>> > Database: Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 -
>> 64bit Production
>> >
>> > The Replicat is slow in terms of no of transactions it is processing
>> and lag is falling behind.
>> >
>> >
>> > <image.png>
>> >
>> > Generated AWR for target database and there is very little activity and
>> the no of sql executions are very low when compared with other replicat in
>> a different environment. The rate at which the replicat is reading the
>> trail file is very slow.
>> >
>> > <image.png>
>> >
>> > A piece of trail file:
>> >
>> > <image.png>
>> >
>> > Please let me know what would be next phase of troubleshooting. Any
>> help is very appreciated.
>> >
>> > Thanks & Regards,
>> > Ravi Teja.
>>
>

--
http://www.freelists.org/webpage/oracle-l
Received on Sun Dec 17 2017 - 23:14:23 CET

Original text of this message