Re: caPutLog Performance

[Matthias Clausen <Matthias.Clausen@desy.de>]

Hi Gary.

Gary Carr wrote:
> Does anyone have any performance data for caPutLogging? How does it 
> effect the IOC? How many logs can you do per second? 
Gongfa Liu ran several tests while he was working at DESY.
I do not recall any statistical performance numbers he took.
Maybe he can comment on this.

In any case there's a built in queue which performs like a throttle.
> I would be interested in any data for the old system (3.13.4), and the 
> new 3.14 caPutLogger.
There are several improvements in 3.14 which will not affect the CPU 
load but rather the correct behavior when you keep track of put to the 
same channel within a given time (keeping track of the min, max values 
when e.g. writing to a channel with a slider). These changes still need 
to be retrofitted into the current release.

The CPU load was never an issue and so we did neither measure nor change 
anything in this respect.
Beware: If you are writing regularly from a program outside the IOC 
through channel access to the IOC - this will trigger caPutLogging :-(

In our case we are forwarding puLog, snlLog, sysLog, alarms ... through 
logMessage processes to our JMS based logging system.

Cheers
Matthias



--
------------------------------------------------------------------------
Matthias Clausen                         Cryogenic Controls Group(MKS-2)
phone:  +49-40-8998-3256                Deutsches Elektronen Synchrotron
fax:    +49-40-8994-3256                                    Notkestr. 85
e-mail: Matthias.Clausen@desy.de                           22607 Hamburg
WWW-MKS2.desy.de                                                 Germany
------------------------------------------------------------------------