[omniORB] Corba slow

[Peter Klotz]

Hello Jingdong

We use omniORB 4.2.0 under RHEL7 x86_64 and experience no issues like you describe them.

However we ran into a performance problem with a non-CORBA network protocol where it turned out that the MTU size change from RHEL6 (16436 bytes) to RHEL7 (65536 bytes) for the loopback interface was the culprit.
This of course only happens if the client and server run on the same machine.

You can change the MTU size temporarily using "ifconfig lo mtu 16384 up" and permanently by adding "MTU=16436" to /etc/sysconfig/network-scripts/ifcfg-lo.

Regards, Peter.

Von: Jingdong Sun [mailto:jindong at us.ibm.com]
Gesendet: Dienstag, 11. November 2014 19:10
An: omniorb-list at omniorb-support.com
Betreff: [omniORB] Corba slow

Hi, there,

I experience some corba slowness when using corba with RH7?

I am using omniORB 4.1.6 with following setup (other config related to thread and connection all take default):
-ORBacceptBiDirectionalGIOP,1,
-ORBofferBiDirectionalGIOP,1,
-ORBthreadPerConnectionPolicy,1,
-ORBthreadPerConnectionLowerLimit,40,
-ORBthreadPerConnectionUpperLimit,50,
-ORBthreadPoolWatchConnection,0,

RedHat 7.0 x86_64 version 2.4.5

The slowness I noticed is when client trying to call to server, some times it hang there for tens of seconds to several minutes before timeout and retry (2nd retry generally work).
For RH 6.x, I did not notice this, or at least not happened this often and noticeable.

Any suggestion to fix this...I feel that some config settings may help here, but I do not know what.

Thanks.
Jingdong Sun
InfoSphere Streams Development
Phone  507 253-5958  (T/L 553-5958)
jindong at us.ibm.com<mailto:jindong at us.ibm.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.omniorb-support.com/pipermail/omniorb-list/attachments/20141112/5e5c4469/attachment-0001.html>