EPICS Home

Experimental Physics and Industrial Control System


 
1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  <20062007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024  Index 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  <20062007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
<== Date ==> <== Thread ==>

Subject: RE: JCA, Gateways and EPICS_CA_ADDR_LIST
From: "Rees, NP \(Nick\)" <[email protected]>
To: "Marty Smith" <[email protected]>
Cc: "EPICS tech-talk" <[email protected]>, "Matej Sekoranja" <[email protected]>, "Yuan, F \(Fajin\)" <[email protected]>
Date: Wed, 6 Dec 2006 13:08:12 -0000
Marty,

Thanks. That works. It is a much better workaround than any of mine.

Cheers,

Nick Rees
Principal Software Engineer           Phone: +44 (0)1235-778430
Diamond Light Source                  Fax:   +44 (0)1235-446713

> -----Original Message-----
> From: Marty Smith [mailto:[email protected]] 
> Sent: 06 December 2006 12:29
> To: Rees, NP (Nick)
> Cc: EPICS tech-talk; Matej Sekoranja; Yuan, F (Fajin)
> Subject: Re: JCA, Gateways and EPICS_CA_ADDR_LIST
> 
> 
> Hi Nick,
> 
> Have you tried using the subnet broadcast IP in the addr_list rather 
> than specific IPs ?
> 
> Marty Smith
> APS Controls Group
> 
> 
> Rees, NP (Nick) wrote:
> > We have just encountered what seems to be a significant bug in the
> > interaction between CAJ and the CAGateways. I am not sure 
> whether this
> > has been reported before, but I assume someone else must 
> have seen it.
> >
> > The problem is:
> >
> >  - If you set 
> com.cosylab.epics.caj.CAJContext.auto_addr_list to TRUE
> >    it will not resolve PV's being serviced by a local 
> gateway, but will
> >    resolve PV's from a normal local IOC.
> >
> >  - If you set auto_addr_list to false and define an 
> explicit addr_list
> >    then it works, but then addr_list has to contain the 
> addresses of all
> >
> >    your IOC's (plus the gateway) - and this is impossible to manage.
> >
> >  - If you set auto_addr_list to true and addr_list to just 
> the gateway,
> >    this would be a workaround in the C environment but 
> doesn't work in
> > the
> >    CAJ environment.
> >
> > Hence, the only way to get PV's from both gateways and 
> local IOC's is
> > via the second option, and that requires long, unmanageable 
> IOC lists
> > that are different on every subnet. 
> >
> > Has this been seen by anyone? I get the same problems with
> > caj-1.0.4+jca-2.1.6 and caj-1.1.2:jca-2.3.0. The caget test 
> program is
> > attached. Any help would be appreciated.
> >
> > Cheers,
> >
> > Nick Rees
> > Principal Software Engineer           Phone: +44 (0)1235-778430
> > Diamond Light Source                  Fax:   +44 (0)1235-446713
> >   
> 
> 


References:
Re: JCA, Gateways and EPICS_CA_ADDR_LIST Marty Smith

Navigate by Date:
Prev: Re: JCA, Gateways and EPICS_CA_ADDR_LIST Marty Smith
Next: Re: wxPython timer and KEK CaPython314 cause core dump Zhong Ren
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  <20062007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024 
Navigate by Thread:
Prev: Re: JCA, Gateways and EPICS_CA_ADDR_LIST Marty Smith
Next: RE: Problems with Channel Access calls on an IOC moving from R3.13.7 -> 3.14.8 Jeff Hill
Index: 1994  1995  1996  1997  1998  1999  2000  2001  2002  2003  2004  2005  <20062007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017  2018  2019  2020  2021  2022  2023  2024