Sandy,
So if you do a manual "alter system archive log current" on the Primary, it
catches up?
I would look at the config parameters to compare the configuration. Could you
share the configuration info (db/server names suitably redacted):
dgmgrl> show configuration verbose
dgmgrl> show database verbose <primary|standby2|standby2>
regards
Neil Chandler
Database Guy, Knows Things.
________________________________
From: Sandra Becker <sbecker6925@xxxxxxxxx>
Sent: 03 January 2019 22:49
To: John Thomas
Cc: Andrew Kerber; Neil Chandler; oracle-l
Subject: Re: One primary with two physical standbys exhibiting different
behavior with regard to lag
My understanding is the network setup is the same between the standbys. I
didn't look at the network right away. :-) I made the changes suggested by
Neil, but I'm still seeing the delay. Before I did a manual log switch, the
delay was over 30 minutes. Not good for this critical production standby.
Sandy
On Thu, Jan 3, 2019 at 3:36 PM John Thomas
<jt2354@xxxxxxxxx<mailto:jt2354@xxxxxxxxx>> wrote:
There's no excessive delay between your primary and the lagging standby is
there? Smaller pipe? Lots of network retries?
Probably the second thing you checked...
Regards,
John
On Thu, 3 Jan 2019 at 22:10, Sandra Becker
<sbecker6925@xxxxxxxxx<mailto:sbecker6925@xxxxxxxxx>> wrote:
Thanks, Andrew. That's one of the first things I checked. It's the same on
both standbys.
Sandy
On Thu, Jan 3, 2019 at 2:47 PM Andrew Kerber
<andrew.kerber@xxxxxxxxx<mailto:andrew.kerber@xxxxxxxxx>> wrote:
Neil most likely spotted the problem. But you should also check to make sure
that the protection mode is the same on both standbys.If the instance that is
behind is using the maximum performance (async) mode it can run a ways behind
the primary.
On Thu, Jan 3, 2019 at 3:36 PM Neil Chandler
<neil_chandler@xxxxxxxxxxx<mailto:neil_chandler@xxxxxxxxxxx>> wrote:
Sandy,
Have you checked the Standby Redo logs? There's a slight (annoying) change in
Oracle 12.1 onwards which means that Standby Redo logs get created with Thread
0 instead of Thread 1 by default (for a single instance database). Redo can
only use Standby Redo when the threads are the same. If this is RAC you need
Standby Redo for each thread - and you must have 1 more Standby Redo than
Online Redo for each thread.
By coincidence, I wrote a blog post about this 10 minutes ago.
https://chandlerdba.com/2019/01/03/data-guard-unexpected-lag/
regards
Neil Chandler
Database Guy. Knows Things.
________________________________
From: oracle-l-bounce@xxxxxxxxxxxxx<mailto:oracle-l-bounce@xxxxxxxxxxxxx>
<oracle-l-bounce@xxxxxxxxxxxxx<mailto:oracle-l-bounce@xxxxxxxxxxxxx>> on behalf
of Sandra Becker <sbecker6925@xxxxxxxxx<mailto:sbecker6925@xxxxxxxxx>>
Sent: 03 January 2019 20:29
To: oracle-l
Subject: One primary with two physical standbys exhibiting different behavior
with regard to lag
Oracle 12.1.0.2
RHEL7
To begin with, I have not worked much at all with standby databases, so my
knowledge is somewhat lacking.
For business reasons, we have a primary database with two physical standbys.
Everything is configured in dgmgrl and enabled. Monitoring with EM13c is
reporting the lag times, so all looks good for basic setup and monitoring. We
seem to have significant lag at times on one of the standbys, as much as 20
minutes. When looking at v$managed_standby, we see the status as
"WAIT_FOR_LOG". The other standby never seems to be more that a few seconds
behind, if at all, and the status is "APPLYING_LOG".
Is this normal? I've been researching, but haven't found an answer yet. I
didn't create or start the standby databases, so I don't have any idea what was
actually done that could be causing this behavior. Any suggestions would be
appreciated.
Thank you,
--
Sandy B.
--
Andrew W. Kerber
'If at first you dont succeed, dont take up skydiving.'
--
Sandy B.
--
Sandy B.