Re: MOXA ARM computer and streamDevice

[Dirk Zimoch <dirk.zimoch@psi.ch>]

Hello Yong,

I am also interested in this issue.

What is the problem in the cross compiling process?

Can I use NFS on the DA66*?

Are there other alternatives (i.e. 19" computers with many RS232/485 
ports)? So far I used a solution with an additional PC providing remote 
tty devices. But I like the idea of running EPICS directly on the 
machine with the ttys.

Dirk

Hu, Yong schrieb:
> Hi Martin,
>
> I played with Moxa DA662(2 more Ethernet ports than DA661) for a while and made it work with base(3.14.10) + asyn(4.10) + StreamDevice(2.4). The cross compiler(linux-xscale_be) source code is from Moxa, but we rebuilt it. The whole cross compiling process takes some efforts. I can send you the static cross-complied executable(base, asyn and stream all included) if you want.
>
> My application is to monitor 1-wire temperature sensors (~20K) for NSLS-2. We need 90 Moxa DA662, that's a big number. For better system maintenance and software development, I'm going to abandon Moxa DA-662 and changing to use 5 powerful and stable servers(multiple cpu cores) to handle all 1-wire stuff. Another reason I don't like Moxa DA66x is that the Linux/Debian inside Moxa has limited tools/commands so that it's painful to use.
>
> Hope you make a good decision.
>
> Yong
>
> > -----Original Message-----
> > From: tech-talk-bounces@aps.anl.gov [mailto:tech-talk-
> > bounces@aps.anl.gov] On Behalf Of Martin Konrad
> > Sent: Friday, October 01, 2010 2:20 PM
> > To: EPICS tech-talk
> > Subject: MOXA ARM computer and streamDevice
> >
> > Hi all,
> > we are thinking about running IOCs for RS232 devices on ARM based data
> > aquisition computers from MOXA (DA-661):
> > http://store.moxa.com/j2bc_ext/doc/prodbase/M20090309140/DA-660-
> > I_Series.pdf
> >
> > Can someone who already works with similar devices tell us if this
> > should work
> > with base + asyn + streamDevice? What tool-chain did you use for cross
> > compiling? Did it work smoothly? What are your experiences regarding
> > time
> > effort for cross compiling? - Of course setting up a standard Linux PC
> > together
> > with a simple device server does the same job. But fewer devices/PCs
> > means
> > less trouble...
> >
> > best regards
> >
> > Martin
> >
> > --
> > Dipl. Phys. Martin Konrad
> > Technische Universität Darmstadt
> > Institut für Kernphysik
> > Schlossgartenstr. 9
> > 64289 Darmstadt
> > Tel: +49-6151-16-5121
> > Fax: +49-6151-16-4321