Nagy László wrote:
Hello,
I need to setup an environment where some users (10 to 20 employees)
will use terminals to run programs. They need to run a few popular
programs: thunderbird, firefox, adobe acrobat, openoffice and gaim. This
site will be a customer service. We decided to reduce the costs by using
Open Source software and cheap terminal computers. This is a good
solution because most of the users will read messages and images on the
screen and they can share the same processor and memory easily. I know
that I can setup cheap computers and use its X server as a terminal for
another central computer. This solution still requires new (or used)
computers. I would like to reduce the costs to the minimum. Here are
some key questions that I could not answer:
- Is there a more cost-effective solution? (Something that I did not
think of)
- How much RAM will I need? Will FireFox Thunderbird and OpenOffice load
shared objects and reduce the overall memory usage? Or should I reserve
256MB of memory for each client?
- Do I need to use gigabit ethernet? Or is it enough to use a normal 100
Mbps wired network? I heard that there can be bandwidth problems when
using many terminals, but I do not have experience.
- Are there any pitfalls that I need to be aware of?
It would be perfect to provide links to some articles or manuals - I do
not need anyone to write detailed instuctions and do my job. I'm asking
for help because the handbook was not very useful in this case. I only
found this:
http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/term.html#TERM-X
It does not help too much, and there is no know-how. I really need to
know what hardware I need to buy.
Remember that the main cost is maintenance, not the hardware.
I think that the way to do it is not dumb terminals in the old sense,
but rather sharing disks, while each terminal runs processes separately
and have lot's of RAM - 1GB.
100Mbps network should be ok, just make sure it's switched (which all
are nowadays), it's only loading the applications that is slow - once
up, there is not much on the network when applications run on the client
and there is plenty of RAM. I would think that more RAM gives better
user experience than faster network.
Some recommends booting off a flashrom, but the disadvantage is
upgrading the base system has to be done on each client.
For example: Buy some mini-itx MB's with 1GB ram. For desktop use,
processor is not important, RAM is. So get some fanless MB's. I have
found that VIA MB's are easy to work with, support pxeboot, see this
site: www.mini-itx.com.
Then you need one file server to allow NFS mount of everything. I sat
down and wrote about it, but I never got through to have a working
diskless with all the bells and whistles, see this article:
www.daemonsecurity.com/pub/pxeboot/
Other sources are the pxe and diskless articles in
www.freebsd.org/doc/en_US.ISO8859-1/articles/
Cheers, Erik
--
Ph: +34.666334818 web: http://www.locolomo.org
X.509 Certificate: http://www.locolomo.org/crt/8D03551FFCE04F0C.crt
Key ID: 69:79:B8:2C:E3:8F:E7:BE:5D:C3:C3:B1:74:62:B8:3F:9F:1F:69:B9
_______________________________________________
freebsd-questions@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
