Re: EPICS archiver

["Eugene T. Yamamoto" <geno@physics.ucla.edu>]

Actually, AR does a core dump. No intentional aborts as far as I can see.
Maybe we just have a REALLY old version of AR.
-Eugene Yamamoto

On Wed, 30 Jun 1999, Brian McAllister wrote:

> 
> > here another question. I am trying to use the archiver for EPICS, but
> > there are several problems
> > If I am using the snapshot mode, it could miss some things happened in
> > between. If I use the monitor mode, it crashed when it reaches the file
> > size that I specified.
> 
> It doesn't crash, it aborts intentionally (due to failure of an "assert()"
> test on the success of a write to the file).
> 
> Not very friendly behavior, especially if you have more than one request
> open, as it won't properly flush and close the others before exiting.
> 
> > And it gets tothat file size too fast. $ variables only take 3 hours to
> > reach 1Mbytes.
> 
> The maximum file size allowed by AR seems to be 100 Mbytes.  This is set in
> the OUTPUT line in the request file.
> 
> Also, look carefully at the data you are archiving.  It may be possible to
> reduce the data rate by increasing deadbands (.ADEL) or using smoothing in
> your database.
> 
> ----
> Brian McAllister                 Application Programmer/Beam Physicist
> mcallister@mit.edu                        MIT-Bates Linear Accelerator
>