[omniORB] Re: Bug? Calls serialization in omni4.0.1

[Dmitry Davidovich]

Thread id tracing discovered next behavior:
long call sequences (20 calls) dispatched to the same thread with only 2
threads in use (on 4 CPUs),
BUT !!!! it's true for relatively short calls only (5 ms). If I make a
longer busy loop (100 ms) all CPUs will be utilised.

Any idea?

Dmitry

> -----Original Message-----
> From: baileyk at schneider.com [mailto:baileyk at schneider.com]
> Sent: Tuesday, 01 July 2003 18:42
> To: Dmitry Davidovich
> Cc: 'omniorb-list at omniorb-support.com'
> Subject: RE: [omniORB] Bug? Calls serialization in omni4.0.1
> 
> 
> 
> I run both WebSphere (v4 and now v5 too) and omniORB on Solaris E10K
> hardware, with system images of up to 12 processors I think.  Sorry I
> missed the detail in your post.
> 
> Some of the calls to the CORBA server in my app run for quite 
> a long time
> (minutes) while others are fast.  I can't say for sure that under load
> WebSphere never opens more than one connection to omniORB, 
> but in testing
> it sure appeared to multiplex all calls on a single 
> connection.  If all
> these calls were serialized it would be a disaster.
> 
> Have you turned tracing on in the omniORB server (with thread 
> tracing) and
> verified that concurrent requests are being handled sequentailly by a
> single thread?  Your first post quoted some CPU utilization 
> numbers which
> suggest such, but are you sure?  Maybe there's something else 
> going on?
> 
> Kendall