JD Edwards EnterpriseOne 9.2, Tools Release 9.2.0.5: UBE Performance Issue

rayrubio

Member
Hello Everyone,

My company is experiencing the issue as described in the following Oracle document:

E1: UBE: UBE Performance Slows After Upgrade to 92 (Doc ID 2177571.1)


https://support.oracle.com/epmos/fa...e=156zmw8195_295&_afrLoop=315805453476589#FIX

The issue is that we are experiencing are that UBEs are taking significantly longer to run in JD Edwards EnterpriseOne 9.2, Tools Release 9.2.0.5. There was a fix released after Tools Release 9.2.0.3, but we are still having the slowness issue after the fix is in place. The fix was adding the following to the jde.ini - CallobjMonitorWorkingDir=0.

Our JD Edwards environment is as follows:

JD Edwards 9.2 Version: 9.2
JD Edwards Tools Release Version: 9.2.0.5
JD Edwards Servers OS: Windows Server 2012 R2 (64-bit)
Database: Microsoft SQL Server 2014
Oracle WebLogic Version: 12.2.1.0.0

Here are examples of applications that are taking a long time to run:

Credit Analysis Refresh, R03B525 - Takes an hour to run in 9.2. In 9.0 takes 3 to 4 minutes.
Pre-Payroll, R07200 - Takes an hour or more to run in 9.2. In 9.0 takes 3 to 4 minutes.


Thank you for any help!
 
Hi rayrubio,

We are experiencing the same issue (also with Tools Release 9.2.0.5). In our situation, it appears that SQL statements are taking milliseconds longer to prep and execute than it did in previous versions. If you have a job that submits millions and millions of SQL statements, the millisecond time difference adds up to hours of extra processing.

With this in mind, I asked a question in a different thread that is based around why statements could be taking longer. There are a handful of great suggestions that you may want to pursue. We just went live with 9.2.0.5 earlier this week so we haven't had a chance to try anything yet. Therefore, we are still experiencing issues with long running jobs. However, perhaps you are in a position where these tips may lead you in the right direction.

http://www.jdelist.com/vb4/showthread.php/52510-SQL-Performance-Difference-(IBMi-vs-SQL-Server)

Good luck!

Chop
 
Hello Everyone,

My company is experiencing the issue as described in the following Oracle document:

E1: UBE: UBE Performance Slows After Upgrade to 92 (Doc ID 2177571.1)


https://support.oracle.com/epmos/fa...e=156zmw8195_295&_afrLoop=315805453476589#FIX

The issue is that we are experiencing are that UBEs are taking significantly longer to run in JD Edwards EnterpriseOne 9.2, Tools Release 9.2.0.5. There was a fix released after Tools Release 9.2.0.3, but we are still having the slowness issue after the fix is in place. The fix was adding the following to the jde.ini - CallobjMonitorWorkingDir=0.

Our JD Edwards environment is as follows:

JD Edwards 9.2 Version: 9.2
JD Edwards Tools Release Version: 9.2.0.5
JD Edwards Servers OS: Windows Server 2012 R2 (64-bit)
Database: Microsoft SQL Server 2014
Oracle WebLogic Version: 12.2.1.0.0

Here are examples of applications that are taking a long time to run:

Credit Analysis Refresh, R03B525 - Takes an hour to run in 9.2. In 9.0 takes 3 to 4 minutes.
Pre-Payroll, R07200 - Takes an hour or more to run in 9.2. In 9.0 takes 3 to 4 minutes.


Thank you for any help!


Same answers I posted here: http://www.jdelist.com/vb4/showthread.php/52510-SQL-Performance-Difference-(IBMi-vs-SQL-Server)

- SQL Index Fragmentation
http://jeffstevenson.karamazovgroup.com/2008/09/determine-index-fragmentation-in-all.html
- Processor Power Management
http://jeffstevenson.karamazovgroup.com/2011/06/windows-2008s-magic-performance-button.html
 
Are you on VMWare? I found the network configurations that fixed our issues. Everything runs as normal now.
 
VMWare 6.0 - Microsoft SQL Server 2014 on Windows Server 2012

Here are the network settings that were configured to fix the performance issue with the Microsoft Windows 2012 Servers running in VMWare. The configurations were completed on all JD Edwards 9.2 servers on 12/01/2016.

The issue is caused by the Receive Segment Coalescing (RSC) feature found on the newer VMWare VMXNET3 NICs on Windows Server 2012.

You can disable RSC globally or on a per interface basis.

Disable RSC globally:

netsh int tcp set global rsc=disabled



Note: This command requires an elevated command prompt.

To verify that RSC is disabled on the virtual machine:

Use the netsh int tcp show global command. The Receive Segment Coalescing State should be set to disabled.


Disable LRO on a VMXNET3 Adapter on a Windows Virtual Machine:

On Windows, the LRO technology is also referred to as Receive Side Coalescing (RSC).

Prerequisites


Verify that the virtual machine runs Windows Server 2012 and later or Windows 8 and later.

Verify that the virtual machine compatibility is ESXi 6.0 and later.

Verify that the version of the VMXNET3 driver installed on the guest operating system is
1.6.6.0 and later.

1
In the Network and Sharing Center of the guest operating system's Control Panel, click the
Name of the network adapter.
A dialog box displays the status of the adapter.
2
Click Properties, and under the VMXNET3 network adapter type, click Configure.
3
On the Advanced tab, set both Recv Segment Coalescing (IPv4) and Recv Segment
Coalescing (IPv6) to Disabled.
4
Click OK.




References:

Enable or Disable LRO on a VMXNET3 Adapter on a Windows Virtual Machine

https://pubs.vmware.com/vsphere-60/...UID-ECC80415-442C-44E9-BA7A-852DDB174B9F.html


VMXMET3 Adapter on Windows Server 2012 with MSSQL Server Bottleneck Problem
https://communities.vmware.com/thread/524842?start=0&tstart=0
After upgrading a virtual machine to hardware version 11, network dependent workloads experience performance degradation (2129176)

https://kb.vmware.com/selfservice/m...nguage=en_US&cmd=displayKC&externalId=2129176

Receive Segment Coalescing (RSC)

https://technet.microsoft.com/en-us/library/hh997024(v=ws.11).aspx


https://kb.vmware.com/selfservice/m...nguage=en_US&cmd=displayKC&externalId=2129176
 
Thanks for sharing Ray. Glad you found the cause.

You're welcome Larry. I hope this fix will help others. It was like finding a needle in a haystack, since it wasn't really a JD Edwards issue. Moreso, it was a Windows 2012 and Microsoft SQL 2014 issue in VMWare.

By the way, there are other standard Virtual Machine performance tweaks that we configure on Windows 2012 servers. The fix I posted above is what fixed this specific issue, but we also do other tweaks for 2012 server in VMWare. If anyone is interested message me or email me at [email protected]. -Ray-
 
So did you end up looking at these two?

We had originally thought that the issue was due to SQL Index Fragmentation, but that was not the case.

We are also aware of the Processor Power Management as well, but that was also not the fix for our issues. We have a JDE EnterpriseOne 9.0 environment running perfectly fine in VMWare. The only difference between our 9.0 environment and our 9.2 environment is that we are running Windows 2012 and Microsoft SQL 2014 on 9.2. The fixes I noted above was specific to VMWare 6, Microsoft Windows Server 2012, and Microsoft SQL Server 2014. As soon as the configurations were complete, our 9.2 performance drastically improved.
 
We're experiencing the same UBE slowness for our 9.2 upgrade as well (9.2.1.1 tools to be precise). Our specs are similar to the OP, except we're running on an Oracle database (12.1.0.2.0).

The Ent/JAS servers are running on VMWare 6.0, except the database which is standalone. Do the VMWare adjustments listed above only apply when running on Microsoft SQL server 2014 database, or could this be a solution for us as well running Oracle?

Thanks.
 
Back
Top