Thanks for the response. You make a lot of sense for sure. I did a test a
few weeks back with one of the designers machines by cabling it directly to
the gigabit switch and it was quite different... The logon/off process was
quite a bit snappier (he runs login and logoff scripts) and I also noticed
that opening and saving files was a lot faster. It was then that I thought
that I may as well do this being as we have the capabilities and other than
my time, it's really a no cost job. We have 1000' of CAT5e cable and the
available switches so are good to go.
Again, I appreciate your response...
Thanks, Brad
"Phillip Windell" <> wrote in message
news:OqS$...
> "Brad Pears" <> wrote in message
> news:...
>
>> Is it really going to make that big of a difference to cable them from
>> the 8 port vs cabling some users direct to the same switch the servers
>> are plugged into??
>
> Gigabit is gigabit is gigabit is gigabit. If everything is running at
> gigabit speed than that is what you are going to get no matter what you
> do. There isn't going to be any additional speed from anywhere. However
> you can slow the speed by running multiple users over the same "backbone"
> cable,...but even then that is only if they are transfering file at
> exactly the same moment. If they are transfering at different moments or
> at only slightly overlapping moments then no difference (good or bad) is
> going to be noticed.
>
> Normally a Backbone runs at a faster speed than the Hosts and it runs from
> specific ports on the switch. For example a 100mbps LAN that uses a
> 1000mbps backbone. This way there could be 10 users hitting the backbone
> at exactly the same moment and it could handle it.
>
> If you have a completely 1000mbps (gigabit) LAN and your Backbone is also
> 1000mbps then you don't even really have a "backbone",...all you have is
> an over-loaded cable connecting the switches together. It would need a
> 10gig Backbone link with proper cabling capable of handling that high of
> speed with 10gig uplink ports in the switches to handle them.
>
> With all that out of the way,...most of the time I run into a question
> like this the person is trying to solve and "imagined" problem that does
> not exist. These file transfers do not take more than a few seconds at
> worst,...and they do not happen constantly. They happen when the user
> opens the file and when they close the file,...and that is it. It is not
> "constant traffic",...the LAN could be completely quiet the rest of the
> time. This might occur a few times an hour as the user "saves" the file
> along the way as the designer works on the file. this is by no means a
> heavey load on the LAN. The difference in speed of opening/saving the
> file between 1gig or 10gig (and maybe even 100mbps) is probably only going
> to be one to three seconds,...that is not going to "hold up" anyone up in
> getting their work done.
>
> I'm not saying you shouldn't strive to make the LAN more efficient,..you
> should,...I'm just saying that you have to keep everything in
> perspective,...focus on what matters,...focus on real measurable tangable
> problems.
>
> --
> Phillip Windell
>
> The views expressed, are my own and not those of my employer, or
> Microsoft,
> or anyone else associated with me, including my cats.
> -----------------------------------------------------
>
>
|