Experimental Physics and
Industrial Control System

"Jeff Hill" <[email protected]> · Sat, 9 Oct 2010 15:24:04 -0600

Hi Patrick,

So I will concur with Ben. This is of course
the classic consumer producer situation. If the 
producer rate exceeds consumer rate no amount of 
buffering will help long term, and robust client side 
software should always concede at least a small 
probability that the server might produce at a rate 
faster than the lower priority server threads in the 
IOC, network, or client side application can handle.

One can greatly reduce occurrence rates of discarded 
intermediate values by reducing CPU loading competing
with the IOC's per client thread, by streamlining
the network layers, and by streamlining the client 
side application. Frequently the issue is something the
client side application does that reduces throughput
like writing too few bytes per IO request to the disk, 
formatting text, and or writing to standard out.

In summary, occurrence rates can be reduced but a
client side application intolerant of situations where 
intermediate events are discarded might not be an
a realistic / general-purpose design.

Jeff
______________________________________________________
Jeffrey O. Hill           Email        [email protected]
LANL MS H820              Voice        505 665 1831
Los Alamos NM 87545 USA   FAX          505 665 5107

Message content: TSPA

> -----Original Message-----
> From: [email protected] [mailto:tech-talk-
> [email protected]] On Behalf Of Patrick Thomas
> Sent: Friday, October 08, 2010 4:43 PM
> To: [email protected]
> Subject: subscription updates
> 
>   Hi,
> 
> The channel access reference manual states that:
> 
> There is unfortunately no field in the protocol allowing the server to
> indicate that an intervening subscription update was discarded.
> 
> Is there a way to ensure that an intervening subscription update will
> not be discarded? Would you have to ensure that each callback executes
> in a certain amount of time, ie, faster that the channel update
> interval?
> 
> Thank you,
> Patrick

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System