We sometimes tend to forget the upgrades can go smoothly too, at least from a
database perspective. :)
Indeed. This was only the first step of the hurdle, this is really a functional
test environment. Next week when we move the Performance test environment, I
expect a lot more fun following..
-Upendra
________________________________
From: Gerald Venzl <gerald.venzl@xxxxxxxxxx>
Sent: Friday, September 15, 2017 5:37 PM
To: nupendra@xxxxxxxxxxx
Cc: Stefan Koehler; xt.and.r@xxxxxxxxx; Oracle-L
Subject: Re: Weird Oracle 12.2 issue .. - FIXED
So there are cases where the network is blamed rightfully!
And I always thought that “it’s the network” is just a lame excuse. :)
Glad to hear that your issue has been solved and that the performance is on-par
or even better than 11g.
We sometimes tend to forget the upgrades can go smoothly too, at least from a
database perspective. :)
Cheers,
---------------
Gerald Venzl | Senior Principal Product Manager
Email: gerald.venzl@xxxxxxxxxx<mailto:gerald.venzl@xxxxxxxxxx> | Phone:
+1.650.633.0085<tel:+16506330085>
Oracle ST & Database Development
400 Oracle Parkway | Redwood Shores | 94065 | USA
On Sep 15, 2017, at 12:40, Upendra nerilla
<nupendra@xxxxxxxxxxx<mailto:nupendra@xxxxxxxxxxx>> wrote:
Thanks everyone for the great feedback and pointers.
It turns out to be something in the underlying VM environment..
When we migrated the DB from 11g to 12c environment, the underlying VM farm was
setup in a different network than where it was.. That traffic ended up going
through a couple more switches and a firewall which added more latency..
Even though the network tests should have caught them they somehow were kept
saying everything was good.😞
Both networks were fine, it is only the communication across the networks which
were the issue.. nevertheless it should have been caught..
This morning when I spoke to someone else the different zoning of the VM came
to light. We migrated the VM to the other zone which fixed most of the issues.
We had to copy a SQL profile from 11g to 12c to address another issue. Looks
like the 12c response times are on-par or better than 11g. Yay!!
Thanks again for everyone in the thread in chiming in and providing valuable
feedback..
-Upendra
________________________________
From: Stefan Koehler <contact@xxxxxxxx<mailto:contact@xxxxxxxx>>
Sent: Friday, September 15, 2017 10:55 AM
To: xt.and.r@xxxxxxxxx<mailto:xt.and.r@xxxxxxxxx>
Cc: Oracle-L; Upendra nerilla
Subject: Re: Weird Oracle 12.2 issue ..
Hi Sayan,
yes, as the LOB pre-fetching i was talking about is something different:
http://docs.oracle.com/database/122/LNOCI/lobs-and-bfile-operations.htm#LNOCI07100
8 LOB and BFILE Operations -
Oracle<http://docs.oracle.com/database/122/LNOCI/lobs-and-bfile-operations.htm#LNOCI07100>
docs.oracle.com<http://docs.oracle.com/>
OCI includes a set of functions for performing operations on large objects
(LOBs) in a database. Persistent LOBs (BLOBs, CLOBs, NCLOBs) are stored in the
database ...
Best Regards
Stefan Koehler
Independent Oracle performance consultant and researcher
Website:
http://www.soocs.de<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.soocs.de&d=DwMGaQ&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=zUfLpQOHlJTnMpyBPavQbdKRMt_LVE0xNvj73TLA4pQ&m=1OJh4zNzHK6460EgqWEbNRPmBecQo3hV4heqAFSG65Y&s=l-A0yMtae6ou8H-Wb6TGPElH6HKGv5R4D1qj-0sXJt8&e=>
soocs - Your reliable partner for top-notch Oracle
...<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.soocs.de_&d=DwMGaQ&c=RoP1YumCXCgaWHvlZYR8PQcxBKCX5YTpkKY057SbK10&r=zUfLpQOHlJTnMpyBPavQbdKRMt_LVE0xNvj73TLA4pQ&m=1OJh4zNzHK6460EgqWEbNRPmBecQo3hV4heqAFSG65Y&s=fU92SS9KWhJ2pbhcrzKrRLnw515nJ7YNtC9QlJGnHOQ&e=>
www.soocs.de<http://www.soocs.de/>
Soocs is an independent Oracle database consultancy that focuses on the Oracle
database technology stack. It is located in Germany and operates throughout the
whole ...
Twitter: @OracleSK
Sayan Malakshinov <xt.and.r@xxxxxxxxx<mailto:xt.and.r@xxxxxxxxx>> hat am 15.
September 2017 um 16:31 geschrieben:
Hi Stefan,
I've just tested LOBPREFETCH parameter in sqlplus 12.2 with the following
query:
SQL> set lobprefet 10000 arrays 1000 autot trace stat;
SQL> select to_clob('1') from dual connect by level<=100;
and it seems it can't reduce extra roundtrips for small lobs:
SQL*Plus 12.2: 302 SQL*Net roundtrips to/from client
SQLcl : 182
SQL*Developer: 11