RE: CA loss package on 1 Hz rate

["Liyu, Andrei" <liyua@ornl.gov>]

Hi,
	I don't use any (sync or async) caget method. I use monitor
method. So I don't lost data like your case. Also I use caget for simple
(=test) program. For normal program monitor is better.
	I also don't care about CPU loading in my case. I control that
and I know when program uses that. It is tied to trigger and I know how
many mc uses BLM thread after trigger. After that CPU is unloaded and
CAS can take everything.

Andrei.

-----Original Message-----
From: Motoki Ooi [mailto:ooi@cens.tokai.jaeri.go.jp] 
Sent: Monday, August 22, 2005 9:08 PM
To: Liyu, Andrei; tech-talk@aps.anl.gov
Subject: Re: CA loss package on 1 Hz rate

Hello

  In my EPICS system, I got similor situation.
 Data is only one Ai (or CALC) record. And there is two PC linux EPICS
IOCs.
 Data is generate on first EPICS IOS and access from second EPICS IOC 
(AI record) to read it (CAget).
 Both EPICS IOC works each SCAN period 40msec  (25Hz).
 Channel Archiver is working on the second PC and archiveing both first 
and second IOCs data.
Archived data from first IOC is almost perfect, but data from second IOC

have some data loss or like that.
I think this data loss due to same reason of Andrei's trouble.
It means, due to the difference of clock frequency of PC (not CPU 
frequency).
For example, first PC may be work at 1.0001 second par 1 second.
And second PC may work at 1.0002 second par 1 second.
In this case, after 400 second, CA cycle of second PC outrun the first 
PCs, and double the archived data.
I think this is the source of the problem.

caution: SCAN period on my EPICS IOCs are set at 0.039 second and works 
at "almost" 40 msec with 1000Hz tick frequency.

Motoki Ooi



Liyu, Andrei wrote:

>Hi,
>
>	SNS BLM system has 2 levels. First level is VME based/vxWorks
>(IOC). Second level is PC/Windows (PC) - similar CAGateway. IOC sends
>data one time per second to PC. Data is arrays (one array (720 floats)
>per detector; up to 32 detectors per IOC). Array has information about
>serial number of current package. PC checks serial number and prints on
>terminal port if data was lost or come twice. Time to time (very
seldom)
>I read both problems.
>	SNS works on 1 Hz now. IOC CPU loading ~ 3-5% and PC CPU loading
>5%. On 60 Hz IOC will have 30% on 60 Hz. PC doesn't depend of
>accelerator rate but depends from clients quantity.
>	I hoped to use CAArchive to find source of problem. But I met
>another problem. Archive lost connection to IOC when IOC/PC connection
>was Ok.
>
>	Does anyone see this problem? Any other suggestions to find the
>source (IOC, network, PC)?
>
>Thanks, Andrei.
>
>
>  
>