structured wiring VS. wireless
Mark
I'm wondering if it's really worth the trouble to retrofit my house with
audio/video/data if there is not an IMMEDIATE need. I currently only
require a few data wires to be pulled. Instead of going all out right now,
should I wait and see what wireless technology advances bring us in 3-5
years? Will residences even be using Cat 6 or equivelent in that time?
Thank in advance,
Mark
Ryan Erickson
How much wire you should install now vs. later depends on a lot of things:
1. How much more is it to have them do it all now (is there a 'visit' cost
as well as the cost to pull the wire).
2. How hard is it to pull in your house?
3. Are the walls drywall or plaster and lath?
4. Is the basement unfinished?
5. Will it be finished in 3-5 years when you need that extra capacity?
6. Would your SO rather have ONE BIG mess now, or ONE MEDIUM mess now and a
bunch of SMALL messes later?
Personally, I would just do it all at once. It isn't much harder to pull
multiple wires than pulling just one, and having the 'possibilities' in the
wall is really where it's at for me. Rather than saying "what wire do I
need to go (find / buy / drill for / install / terminate) to do that?", if
they're already in the wall, you can just say "what do I need to plug into
what"... Much more convenient.
Of course, I'm biased, since I just built a new house, and pulled TONS of
wire (12K+ feet of CAT5 / CAT5E, 1500 feet of RG-6QS, speaker wire, alarm
wire, HA wire, not to mention the 110/220 wiring).
As far as wireless, my guess is that it will always be more expensive,
slower, and much fussier.
Here's an early wiring picture from the house:
http://erickson.myip.org/pictures/newhome/windows/WireMess1.jpg
Good Luck!
Ryan Erickson
use...@erickson.myip.org
http://erickson.myip.org
P.S. Even if you have them do "all" the wires now, you'll have some you've
forgotten and / or need added later... It's the curse of HA...
"Mark" <ma...@nocontact.com> wrote in message
news:Pluj9.38485$gA4.6093@sccrnsc02...
J. Clarke
The cost of pulling wire from one point to another isn't that much
greater if you're pulling 4 wires than if you're pulling 1. Pulling to
different points is a different story.
Talk to your installer and ask him for a quote on just the minimum you
think you need, 2 coax and 2 CAT5e/6 to those points in addition to the
minimum, and 2 coax and 2 CAT5e/6 to two separate points in each room,
in addition to the minimum you think you need. Once you have the actual
numbers you'll be in a better position to decide what you want to do.
Wireless might improve or it might get worse. There are a lot of
services sharing that band already, and the number will increase, which
can't help but ultimately result in interference even with spread-
spectrum techniques. Personally I'd be quite surprised to see an
economical and reliable 1 Gb/sec residential wireless service available
in my lifetime (mainly because I don't think that the FCC is going to
make enough bandwidth available for unlicensed residential services to
allow that), but Gigabit Ethernet is already cheap enough that if I
wanted it badly enough I could install it today, and it's going to be a
lot cheaper in a couple of years.
In the 5 year timeframe, it's unlikely that CAT5e will be superseded by
a different technology--5e will carry a billion bits/sec reliably, and
there aren't any residential services that I know of that can fill up
even a 100 Mb/sec pipe. The next step beyond gigabit is 10 gig and
there are currently no plans to try to run that on copper, at least not
for general purpose use--there is discussion of a short-distance version
that runs on copper for internal use in corporate data centers and the
like. So some kind of fibre would be the next step and until the dust
settles on the 10 gig standard and the price has come down to the point
that you're actually ready to buy hardware it would be premature to pull
fibre for that purpose.
> Thank in advance,
>
> Mark
>
>
>
--
--
--John
Reply to jclarke at ae tee tee global dot net
(used to be jclarke at eye bee em dot net)
Mark Lloyd
Currently, wireless is much less reliable and has more problems.
--
93 days until the winter solstice celebration
Mark Lloyd
http://go.to/notstupid
http://notstupid.laughingsquid.com
"I don't believe in an afterlife, so I don't have to
spend my whole life fearing hell, or fearing heaven
even more. For whatever the tortures of hell, I
think the boredom of heaven would be even worse."
-- Isaac Asimov
Mark Lloyd
wrote:
I don't expect gigabit ethernet to help much (over 100Mbit) in a small
network, unless it's very heavily used.
>In the 5 year timeframe, it's unlikely that CAT5e will be superseded by
>a different technology--5e will carry a billion bits/sec reliably, and
>there aren't any residential services that I know of that can fill up
>even a 100 Mb/sec pipe. The next step beyond gigabit is 10 gig and
>there are currently no plans to try to run that on copper, at least not
>for general purpose use--there is discussion of a short-distance version
>that runs on copper for internal use in corporate data centers and the
>like. So some kind of fibre would be the next step and until the dust
>settles on the 10 gig standard and the price has come down to the point
>that you're actually ready to buy hardware it would be premature to pull
>fibre for that purpose.
>
>> Thank in advance,
>>
>> Mark
>>
>>
>>
>
>--
--
PlasmaJohn
> In the 5 year timeframe, it's unlikely that CAT5e will be superseded by
> a different technology--5e will carry a billion bits/sec reliably, and
> there aren't any residential services that I know of that can fill up
> even a 100 Mb/sec pipe.
iSCSI (scsi over ip).
It looks like a disk controller but is just an ethernet card + controller.
There are even soft drivers for clientside use and a slim possibiliy of BIOS
support.
With households acquiring multiple computers, and storage costs falling thru
the floor, a central storage server makes sense. Much easier to build and
manage a single large, hotswappable RAID + backup machine than add those
capabilities to 3 or 4 user machines. :)
PJ
Mark Lloyd
Is that much better than installing drives on a networked computer and
sharing them?
J. Clarke
mll...@godmail.com says...
Not right now. 5 years down the road might be a different story.
Streaming HD video will probably be fairly common by then.
> >In the 5 year timeframe, it's unlikely that CAT5e will be superseded by
> >a different technology--5e will carry a billion bits/sec reliably, and
> >there aren't any residential services that I know of that can fill up
> >even a 100 Mb/sec pipe. The next step beyond gigabit is 10 gig and
> >there are currently no plans to try to run that on copper, at least not
> >for general purpose use--there is discussion of a short-distance version
> >that runs on copper for internal use in corporate data centers and the
> >like. So some kind of fibre would be the next step and until the dust
> >settles on the 10 gig standard and the price has come down to the point
> >that you're actually ready to buy hardware it would be premature to pull
> >fibre for that purpose.
> >
> >> Thank in advance,
> >>
> >> Mark
> >>
> >>
> >>
> >
> >--
>
> --
> 93 days until the winter solstice celebration
>
> Mark Lloyd
> http://go.to/notstupid
> http://notstupid.laughingsquid.com
>
> "I don't believe in an afterlife, so I don't have to
> spend my whole life fearing hell, or fearing heaven
> even more. For whatever the tortures of hell, I
> think the boredom of heaven would be even worse."
> -- Isaac Asimov
>
--
J. Clarke
virus...@projectplasma.com says...
Now, how does this make TCP/IP run faster?
Alan McKay
> > With households acquiring multiple computers, and storage costs falling
thru
> > the floor, a central storage server makes sense. Much easier to build
and
> > manage a single large, hotswappable RAID + backup machine than add those
> > capabilities to 3 or 4 user machines. :)
Actually, central storage makes sense when storage costs are high.
When they are cheap, what makes most sense is to have the HD
in every PC because they are a dime a dozen. The cost of all your
fancy iSCSI SAN and associated equipment is going to be way
more than just buying a HD for each PC.
Don't get me wrong because SANs (Storage Area Networks) are
pretty cool. I just don't see them going into the home any time
soon. In fact, the Metro Optical unit of the company I work for
doesn't even see them going into the average small business
any time soon.
cheers,
-Alan
Alan McKay
> Not right now. 5 years down the road might be a different story.
> Streaming HD video will probably be fairly common by then.
And 100Mbit switched on copper will carry it just fine, thanks.
And GigE will run over copper.
I missed the first of the thread, but if it involved dark fiber, I
consider that to be a big risk, too. To be honest I do not know
how much fiber has changed in the last 5 years, but if it has changed
at all, you can bet it will again in the next 5. Will today's fiber
be good enough for the optical network 5 years down the road?
I dunno. But if I were considering dark fiber I'd be doing some
research on it ...
cheers,
-Alan
Anthony Matonak
...
> Actually, central storage makes sense when storage costs are high.
> When they are cheap, what makes most sense is to have the HD
> in every PC because they are a dime a dozen. The cost of all your
> fancy iSCSI SAN and associated equipment is going to be way
> more than just buying a HD for each PC.
Having centralized storage is good for some things. For instance,
a server which stores various media files, music, video, books, etc.
These files could be accessed anywhere on your network. Voicemail
and security stuff probably makes more sense on a single server than
over a distributed network as well.
Anthony
The.Central.Scru...@invalid.pobox.com
>> > With households acquiring multiple computers, and storage costs falling
>thru
>> > the floor, a central storage server makes sense. Much easier to build
>and
>> > manage a single large, hotswappable RAID + backup machine than add those
>> > capabilities to 3 or 4 user machines. :)
>Actually, central storage makes sense when storage costs are high.
>When they are cheap, what makes most sense is to have the HD
>in every PC because they are a dime a dozen. The cost of all your
>fancy iSCSI SAN and associated equipment is going to be way
>more than just buying a HD for each PC.
It makes to centralize your work files so that they're accessable everywhere
and are backed up.
It doesn't make sense to centralize your apps unless you have to pay for each
copy.
Neil Cherry
>On Mon, 23 Sep 2002 11:42:06 -0400, J. Clarke <nos...@nospam.invalid>
>wrote:
>
>>In article <Pluj9.38485$gA4.6093@sccrnsc02>, ma...@nocontact.com says...
>>> Hi all,
>>>
>>> I'm wondering if it's really worth the trouble to retrofit my house with
>>> audio/video/data if there is not an IMMEDIATE need.
<Snip>
>>Wireless might improve or it might get worse.
<Snip>
I think it will improve but I don't kow about it the security aspects
of it. So far I'm not very trusting of wireless.
>I don't expect gigabit ethernet to help much (over 100Mbit) in a small
>network, unless it's very heavily used.
While I don't see a full gig-e as been a necessity I think gig-e will
be popular if and only if ethernet switching becomes cheap and lots of
devices get an ethernet interface. I kind of see a use for ethernet in
data sharing (obviously with the I'net), audio and video switching and
control (like passing IR). A single MPEG 2 fee is up to 40Mb, picture
in pciture could take up to another 40Mb with TCP/IP overhead and maybe
VLANs I can see a need for more than 100Mb. So the next jump is
gig-e. Now having said that this is still wishful thinking and I've
not seen a whole lot of organization by the various audio and video
groups to take advantage of this 'wonderful idea'. Heck we really
haven't see much in the way of HA really. I'd like to think that this
will change in the next 3 to 5 years but that's what I said in 2000.
BTW, I realize that the above devices would need a pretty good deal of
CPU power to push this technology. Gig-e alone requires a decent
processor to handle the transfer rates, while MPEG 2 would also
require it's own processor, RAM and flash would be pretty hefty. Hmm,
maybe 3-5 is really too 'pie-in-the-sky'.
--
Linux Home Automation Neil Cherry nch...@comcast.net
http://mywebpages.comcast.net/ncherry/ (Text only)
http://linuxha.sourceforge.net/ (SourceForge)
http://hcs.sourceforge.net/ (HCS II)
Neil Cherry
>wrote:
>
>>In article <Pluj9.38485$gA4.6093@sccrnsc02>, ma...@nocontact.com says...
>>> Hi all,
>>>
>>> I'm wondering if it's really worth the trouble to retrofit my house with
>>> audio/video/data if there is not an IMMEDIATE need.
<Snip>
>>Wireless might improve or it might get worse.
<Snip>
I think it will improve but I don't kow about it the security aspects
of it. So far I'm not very trusting of wireless.
>I don't expect gigabit ethernet to help much (over 100Mbit) in a small
>network, unless it's very heavily used.
While I don't see a full gig-e as been a necessity I think gig-e will
J. Clarke
>
> > Not right now. 5 years down the road might be a different story.
> > Streaming HD video will probably be fairly common by then.
>
> And 100Mbit switched on copper will carry it just fine, thanks.
HD to multiple stations with different programming? Don't be too sure.
It's 12 analog MHz per channel.
> And GigE will run over copper.
Yes it will.
> I missed the first of the thread, but if it involved dark fiber, I
> consider that to be a big risk, too. To be honest I do not know
> how much fiber has changed in the last 5 years, but if it has changed
> at all, you can bet it will again in the next 5. Will today's fiber
> be good enough for the optical network 5 years down the road?
> I dunno. But if I were considering dark fiber I'd be doing some
> research on it ...
The question I was addressing was in regard to the future of wireless,
as one might have concluded from the title of the thread. Didn't have
anything at all to do with fiber.
>
> cheers,
> -Alan
J. Clarke
But one does not need a SAN for any of those. They can be done right
now today with the features built into just about every commercially
available, GPLed, or BSDLed OS available that will run on a PC.
J. Clarke
n...@bgp558256bgs.ewndsr01.nj.comcast.net says...
> On 23 Sep 2002 12:37:17 -0500, Mark Lloyd wrote:
> >On Mon, 23 Sep 2002 11:42:06 -0400, J. Clarke <nos...@nospam.invalid>
> >wrote:
> >
> >>In article <Pluj9.38485$gA4.6093@sccrnsc02>, ma...@nocontact.com says...
> >>> Hi all,
> >>>
> >>> I'm wondering if it's really worth the trouble to retrofit my house with
> >>> audio/video/data if there is not an IMMEDIATE need.
>
> <Snip>
>
> >>Wireless might improve or it might get worse.
>
> <Snip>
>
> I think it will improve but I don't kow about it the security aspects
> of it. So far I'm not very trusting of wireless.
>
> >I don't expect gigabit ethernet to help much (over 100Mbit) in a small
> >network, unless it's very heavily used.
>
> While I don't see a full gig-e as been a necessity I think gig-e will
> be popular if and only if ethernet switching becomes cheap
Ethernet switching "became cheap" some time ago. The cost differential
between a hub and a switch is minimal.
> and lots of
> devices get an ethernet interface. I kind of see a use for ethernet in
> data sharing (obviously with the I'net), audio and video switching and
> control (like passing IR). A single MPEG 2 fee is up to 40Mb, picture
> in pciture could take up to another 40Mb with TCP/IP overhead and maybe
> VLANs I can see a need for more than 100Mb. So the next jump is
> gig-e. Now having said that this is still wishful thinking and I've
> not seen a whole lot of organization by the various audio and video
> groups to take advantage of this 'wonderful idea'. Heck we really
> haven't see much in the way of HA really. I'd like to think that this
> will change in the next 3 to 5 years but that's what I said in 2000.
>
> BTW, I realize that the above devices would need a pretty good deal of
> CPU power to push this technology. Gig-e alone requires a decent
> processor to handle the transfer rates, while MPEG 2 would also
> require it's own processor, RAM and flash would be pretty hefty. Hmm,
> maybe 3-5 is really too 'pie-in-the-sky'.
>
>
--
Neil Cherry
>wrote:
>
>>In article <Pluj9.38485$gA4.6093@sccrnsc02>, ma...@nocontact.com says...
>>> Hi all,
>>>
>>> I'm wondering if it's really worth the trouble to retrofit my house with
>>> audio/video/data if there is not an IMMEDIATE need.
<Snip>
>>Wireless might improve or it might get worse.
<Snip>
I think it will improve but I don't kow about it the security aspects
of it. So far I'm not very trusting of wireless.
>I don't expect gigabit ethernet to help much (over 100Mbit) in a small
>network, unless it's very heavily used.
While I don't see a full gig-e as been a necessity I think gig-e will
be popular if and only if ethernet switching becomes cheap and lots of
devices get an ethernet interface. I kind of see a use for ethernet in
data sharing (obviously with the I'net), audio and video switching and
control (like passing IR). A single MPEG 2 fee is up to 40Mb, picture
in pciture could take up to another 40Mb with TCP/IP overhead and maybe
VLANs I can see a need for more than 100Mb. So the next jump is
gig-e. Now having said that this is still wishful thinking and I've
not seen a whole lot of organization by the various audio and video
groups to take advantage of this 'wonderful idea'. Heck we really
haven't see much in the way of HA really. I'd like to think that this
will change in the next 3 to 5 years but that's what I said in 2000.
BTW, I realize that the above devices would need a pretty good deal of
CPU power to push this technology. Gig-e alone requires a decent
processor to handle the transfer rates, while MPEG 2 would also
require it's own processor, RAM and flash would be pretty hefty. Hmm,
maybe 3-5 is really too 'pie-in-the-sky'.
--
Neil Cherry
>wrote:
>
>>In article <Pluj9.38485$gA4.6093@sccrnsc02>, ma...@nocontact.com says...
>>> Hi all,
>>>
>>> I'm wondering if it's really worth the trouble to retrofit my house with
>>> audio/video/data if there is not an IMMEDIATE need.
<Snip>
>>Wireless might improve or it might get worse.
<Snip>
I think it will improve but I don't kow about it the security aspects
of it. So far I'm not very trusting of wireless.
>I don't expect gigabit ethernet to help much (over 100Mbit) in a small
>network, unless it's very heavily used.
While I don't see a full gig-e as been a necessity I think gig-e will
Neil Cherry
>In article <slrnaovhr...@bgp558256bgs.ewndsr01.nj.comcast.net>,
>n...@bgp558256bgs.ewndsr01.nj.comcast.net says...
>> On 23 Sep 2002 12:37:17 -0500, Mark Lloyd wrote:
>> >On Mon, 23 Sep 2002 11:42:06 -0400, J. Clarke <nos...@nospam.invalid>
>> >wrote:
>> >
>> >>In article <Pluj9.38485$gA4.6093@sccrnsc02>, ma...@nocontact.com says...
>> >>> Hi all,
>> >>>
>> >>> I'm wondering if it's really worth the trouble to retrofit my house with
>> >>> audio/video/data if there is not an IMMEDIATE need.
>>
>> <Snip>
>>
>> >>Wireless might improve or it might get worse.
>>
>> <Snip>
>>
>> I think it will improve but I don't kow about it the security aspects
>> of it. So far I'm not very trusting of wireless.
>>
>> >I don't expect gigabit ethernet to help much (over 100Mbit) in a small
>> >network, unless it's very heavily used.
>>
>> While I don't see a full gig-e as been a necessity I think gig-e will
>> be popular if and only if ethernet switching becomes cheap
>
>Ethernet switching "became cheap" some time ago. The cost differential
>between a hub and a switch is minimal.
Sorry, I should have been more specific with my words. I was refering
to Gig-e switches, 10/100Mb switches are the norm though I really wish
I could turn off the switch on a port for proper sniffing on the low
end models (you get what you pay for).
Also, sorry about the triple post. The server replied that it failed
to write the file so I tried 3 times. Guess it lie to me.
PlasmaJohn
> >On Mon, 23 Sep 2002, J. Clarke wrote:
> >
> >> there aren't any residential services that I know of that can fill up
> >> even a 100 Mb/sec pipe.
First, I apologize for taking the thread off-topic, but this is the fragment I
was responding to. I was semi-facetious.
> >iSCSI (scsi over ip).
> Is that much better than installing drives on a networked computer and
> sharing them?
Yep, to subvert certain MS braindamage: you cannot network mount the C: drive.
This is true on most of their platforms, and where it's possible (NT?) it's a
right pain to set up and pretty much limited to an MS proprietary protocol.
What's so special about the C: drive? Every application on the planet wants
to install there and most of them get very cranky if they don't. This is
particularly true of childrens' games.
iSCSI looks like a disk controller and quacks like one too. Windows can't
tell the difference.
PJ
PlasmaJohn
> Actually, central storage makes sense when storage costs are high.
> When they are cheap, what makes most sense is to have the HD
> in every PC because they are a dime a dozen.
Yeah, you sure get what you pay for too. IDE drives are *flaky*. They
fail. Mostly without warning. Does each machine in the house trully need
an 80Gb drive? or even a 40Gb drive?
> The cost of all your fancy iSCSI SAN and associated equipment is going to
> be way more than just buying a HD for each PC.
Sure if you buy an adapter for each machine. You don't need that anyway
(perhaps on the server...). There are soft drivers available today for Win2k
and about every *nix out there. IIRC WinXP will get a MS blessed version RSN.
Ok, so a 100Mbit LAN will run a little poky compared to a GigE network, but
that's "good enough" for a home or a small (3-4) workgroup.
> Don't get me wrong because SANs (Storage Area Networks) are
> pretty cool. I just don't see them going into the home any time
> soon. In fact, the Metro Optical unit of the company I work for
> doesn't even see them going into the average small business
> any time soon.
Perhaps they, like the rest of the SAN industry are focused on the high-end
server farms. The 'experts' have consistently underestimated the home
and SOHO markets [ ... need no more than 3 computers ... need no more than
640k ... need a LAN ... need a firewall ... etc. ].
No, you've got a point. The technology is new enough and raw enough that
the average household would care about it (or even know about it) until it
becomes brainless (read user friendly) and cheap enough. Hmmm.
PJ
john...@yahoo.com
>copy.
why not?
The.Central.Scru...@invalid.pobox.com
>why not?
Because hard drives are dirt cheap and networks are relatively slow.
I'd rather have the apps locally loaded so I didn't have to wait for them to be
sent over the net.
Malcolm W
The Centralized Storage.Scrutinizer wrote in message
Yup. The only drawback is having is making sure the versions are up to date
on the individual workstations.
I sort of miss that brief period (early 90's) where everything was small
enough that it could be run entirely from the network.
In theory anyway. I don't miss working in DOS.
Deborah and Michael Scott
> fail. Mostly without warning. Does each machine in the house trully need
> an 80Gb drive? or even a 40Gb drive?
>
> > The cost of all your fancy iSCSI SAN and associated equipment is going
to
> > be way more than just buying a HD for each PC.
>
> Sure if you buy an adapter for each machine. You don't need that anyway
> (perhaps on the server...). There are soft drivers available today for
Win2k
> and about every *nix out there. IIRC WinXP will get a MS blessed version
RSN.
>
> Ok, so a 100Mbit LAN will run a little poky compared to a GigE network,
but
> that's "good enough" for a home or a small (3-4) workgroup.
Let's assume that you are right when you say that IDE drives are flaky- ( I
have seen both SCSI and EIDE drives be "flaky")- again- we are talking about
a home network. IDE drives are "good enough".
J. Clarke
cha87...@projectplasma.com says...
> On 23 Sep 2002, Mark Lloyd wrote:
>
> > >On Mon, 23 Sep 2002, J. Clarke wrote:
> > >
> > >> there aren't any residential services that I know of that can fill up
> > >> even a 100 Mb/sec pipe.
>
> First, I apologize for taking the thread off-topic, but this is the fragment I
> was responding to. I was semi-facetious.
>
> > >iSCSI (scsi over ip).
>
> > Is that much better than installing drives on a networked computer and
> > sharing them?
>
> Yep, to subvert certain MS braindamage: you cannot network mount the C: drive.
> This is true on most of their platforms, and where it's possible (NT?) it's a
> right pain to set up and pretty much limited to an MS proprietary protocol.
I'm sorry, but I don't see your point. Are you suggesting that instead
of installing an OS on a local $65 drive one instead installs on a
central server and runs across the network? WHY?
That model was abandoned for most purposes a very long time ago with
good reason.
> What's so special about the C: drive? Every application on the planet wants
> to install there and most of them get very cranky if they don't. This is
> particularly true of childrens' games.
Uh, I'm running a Win2K machine dual-boot with 98, with 98 on C and 2K
on D and I'm having no problem at all installing apps on D. But that's
neither here nor there. With Windows you need a local drive regardless.
Even with gigabit trying to run the swap file off of a server is a
losing proposition, and XP/2K will _not_ run without a swap file.
> iSCSI looks like a disk controller and quacks like one too. Windows can't
> tell the difference.
But I still don't understand why anybody would want to do this in a
residential environment. Especially when one notes that the
manufacturers of the interfaces have names like QLogic and Emulex, which
have never been synonymous with "cheap".
> PJ
J. Clarke
n...@bgp558256bgs.ewndsr01.nj.comcast.net says...
Gigabit is getting there. D-Link is selling a 4-port unmanaged gigabit
switch for a street price of around $260.
> Also, sorry about the triple post. The server replied that it failed
> to write the file so I tried 3 times. Guess it lie to me.
>
>
--
J. Clarke
>
> > Actually, central storage makes sense when storage costs are high.
> > When they are cheap, what makes most sense is to have the HD
> > in every PC because they are a dime a dozen.
>
> Yeah, you sure get what you pay for too. IDE drives are *flaky*. They
> fail.
And enterprise class fibre channel drives don't ever fail? You have to
be a newbie.
> Mostly without warning. Does each machine in the house trully need
> an 80Gb drive? or even a 40Gb drive?
When the smallest drive you can buy is 20 gig and it costs 68 bucks
"need" doesn't enter into it.
> > The cost of all your fancy iSCSI SAN and associated equipment is going to
> > be way more than just buying a HD for each PC.
>
> Sure if you buy an adapter for each machine. You don't need that anyway
> (perhaps on the server...). There are soft drivers available today for Win2k
> and about every *nix out there. IIRC WinXP will get a MS blessed version RSN.
Now let's see, earlier you were talking about network-mounting the C
drive as being the incentive to use this technology. Now you're talking
about "soft drivers". Well the soft drivers don't work until the OS has
booted, so you have to install it locally anyway. So you still can't
remap the C drive, and you still have to run the OS from a local drive
and if you're doing that then this brings absolutely nothing to the
table that an ordinary LAN doesn't bring other than increased complexity
at the server end.
Basically all you're accomplishing is to increase the overhead.
> Ok, so a 100Mbit LAN will run a little poky compared to a GigE network, but
> that's "good enough" for a home or a small (3-4) workgroup.
>
> > Don't get me wrong because SANs (Storage Area Networks) are
> > pretty cool. I just don't see them going into the home any time
> > soon. In fact, the Metro Optical unit of the company I work for
> > doesn't even see them going into the average small business
> > any time soon.
>
> Perhaps they, like the rest of the SAN industry are focused on the high-end
> server farms.
Yes, they are. Because those are the only outfits that need what they
provide.
What you are proposing can be done today for a reasonable price with
used fibre-channel equipment. How many home-automation hobbyists are
actually doing it?
> The 'experts' have consistently underestimated the home
> and SOHO markets [ ... need no more than 3 computers ... need no more than
> 640k ... need a LAN ... need a firewall ... etc. ].
>
> No, you've got a point. The technology is new enough and raw enough that
> the average household would care about it (or even know about it) until it
> becomes brainless (read user friendly) and cheap enough. Hmmm.
>
> PJ
>
>
>
--
Neil Cherry
>In article <slrnaovmf...@bgp558256bgs.ewndsr01.nj.comcast.net>,
>n...@bgp558256bgs.ewndsr01.nj.comcast.net says...
>> Sorry, I should have been more specific with my words. I was refering
>> to Gig-e switches, 10/100Mb switches are the norm though I really wish
>> I could turn off the switch on a port for proper sniffing on the low
>> end models (you get what you pay for).
>
>Gigabit is getting there. D-Link is selling a 4-port unmanaged gigabit
>switch for a street price of around $260.
Considering that the Gig-e cards were ~$400 6-12 months ago that's
pretty good.
For those of you who are wondering what this has to do with HA, just
remember that HA is a lot more than monitoring and controlling
lights. Add audio and video and bandwidth usage can jump.
Strange how within the last few years I've seen customers clammering
for more corporate bandwidth. I've watched OC-3's become petite pipes
and customer hub sites requesting OC-12. That would mean that the
shared private network service networks need much bigger pipes (OC-48
& OC-192). This during a very bad economy. I really can't imagine what
the future holds for bandwidth though I know that there's work underway
to build Gig-e WAN interfaces (Gig-e == 1000 Mb, OC-12 = 625 Mb). I
know the router/switch vendors are having fun trying to keep up. Of
course these requests come from very large companies but with the cost
of a DSL/Broadband pipe becoming cheaper than a business pots line
business's are beginning to want bigger pipes to support more data
applications.
Now if we can only work on security, these bigger pipes can lead to
jumbo attacks (outside shared private VPN's).
Malcolm W
"Neil Cherry" <n...@bgp558256bgs.ewndsr01.nj.comcast.net> wrote in message
[snip]
> Now if we can only work on security, these bigger pipes can lead to
> jumbo attacks (outside shared private VPN's).
If those pipes get much bigger, we'll have to start worrying about about
burglers crawling through 'em.
Did they rename T3 to DS-3 to OC-3?
What's the bandwidth for OC-48 and OC-192?
1
> Did they rename T3 to DS-3 to OC-3?
> What's the bandwidth for OC-48 and OC-192?
T3 and DS-3 mean pretty much the same thing, although DS3 is the more
often used term. DS-3 interfaces are typically a pair of BNC coax
connectors. OC(x) are optical interfaces (Optical Connector), although
I'm not 100% sure that is exactly what the OC stands for, but it's close.
An OC-3 is 155Mbps. The OC's based on multiples of
51.85Mbps (an OC-1, which really doesn't exist in common practice).
So, an OC-48 is 2488Mbps (2.48Gbps) and OC-192 is 9955Mbps (9.9Gbps)
By contrast, a T1 is 1.544Mbs, a T2 (yes, such a thing exists, but is
rarely, if ever, actually seen in the wild) is ~4.6Mbps (IIRC) and a
T3/DS3 is ~45Mbps.
Neil Cherry
>"Malcolm W" <wari...@sunsplinkam.net> is rumored to have said:
>
>> Did they rename T3 to DS-3 to OC-3?
>> What's the bandwidth for OC-48 and OC-192?
Sorry, I abuse the T3 and DS3 monicers. Strange how I can consider a
DS3 to be puny (I work on business network services).
>T3 and DS-3 mean pretty much the same thing, although DS3 is the more
>often used term. DS-3 interfaces are typically a pair of BNC coax
>connectors. OC(x) are optical interfaces (Optical Connector), although
>I'm not 100% sure that is exactly what the OC stands for, but it's close.
>An OC-3 is 155Mbps. The OC's based on multiples of
>51.85Mbps (an OC-1, which really doesn't exist in common practice).
>So, an OC-48 is 2488Mbps (2.48Gbps) and OC-192 is 9955Mbps (9.9Gbps)
Nice big pipes delivered to your door. :-} I don't think my Cisco CGS
or 2514 can quite handle that. ;-)
>By contrast, a T1 is 1.544Mbs, a T2 (yes, such a thing exists, but is
>rarely, if ever, actually seen in the wild) is ~4.6Mbps (IIRC) and a
>T3/DS3 is ~45Mbps.
I think a T2 is ~6M (96 channels), I think the only people to see
those are CO techs and maybe a few of the lab folks.
Compare these numbers to DSL with ~384K/128K (is that correct?) and
cable with ~1.5M/128K (at least mine is).
Alan McKay
> connectors. OC(x) are optical interfaces (Optical Connector), although
> I'm not 100% sure that is exactly what the OC stands for, but it's close.
Optical Carrier I always thought. But I don't work in Optical ;-)
cheers,
-alan
Alan McKay
> Nice big pipes delivered to your door. :-} I don't think my Cisco CGS
> or 2514 can quite handle that. ;-)
Not too many people have one of these at their door.
> Compare these numbers to DSL with ~384K/128K (is that correct?) and
> cable with ~1.5M/128K (at least mine is).
Around here DSL generally comes in 1Meg and 3Meg flavours.
1Meg == 1.2Mbit down / 160Kbit up
3Meg == 3.5Mbit down / 800Kbit up
That's residential, done with PPPoE so you lose about 10%
with protocol overhead. My 3Meg connection costs
$50 CDN per month.
Business connections you can get 6Mbit down / 1Mbit up,
as well as 2.3Mbit down / 2.3 up, which is what I'd really like
to have since 3Mbit down is overkill for me, I only want it
for the 800Kbit up that comes with it. But business
connections are damned expensive. ($175 / month
for the 2.3/2.3)
cheers,
-Alan
Dave Houston
>I think a T2 is ~6M (96 channels), I think the only people to see
>those are CO techs and maybe a few of the lab folks.
>
>Compare these numbers to DSL with ~384K/128K (is that correct?) and
>cable with ~1.5M/128K (at least mine is).
My ADSL is 768 Kbps download / 384 Kbps upload. There are some faster
offerings. I believe my son's company has about twice my speed. VDSL can be
more than 10 times the speed of ADSL.
But, the telcos are not likely to be eager to replace existing hardware
since few face any real competition for broadband.
And your cable speed deteriorates with the number of users whereas ADSL is a
one-to-one thing.
---
http://www.laser.com/dhouston/bx24-pcb.htm
Alan McKay
> And your cable speed deteriorates with the number of users whereas ADSL is
a
> one-to-one thing.
This can be misleading. Yes, it's true, but if your DSL provider doesn't
have
a big enough pipe to the internet then your bottleneck simply moves from
the local neighbourhood to the ISP's machine room.
cheers,
-Alan
PlasmaJohn
> And enterprise class fibre channel drives don't ever fail?
"Enterprise" class hardware has quite a bit of redundancy, something I want
for each family member's computer, but adding a RAID to each machine
is a bit expensive.
> You have to be a newbie.
Your dazzling use of hominem-fu has blinded me. I cede the argument.
Neil Cherry
>> Nice big pipes delivered to your door. :-} I don't think my Cisco CGS
>> or 2514 can quite handle that. ;-)
>Not too many people have one of these at their door.
Yeah, it comes from using them in design of commercial networks. You
get spoiled by the tools.
>> Compare these numbers to DSL with ~384K/128K (is that correct?) and
>> cable with ~1.5M/128K (at least mine is).
>Around here DSL generally comes in 1Meg and 3Meg flavours.
>1Meg == 1.2Mbit down / 160Kbit up
>3Meg == 3.5Mbit down / 800Kbit up
>That's residential, done with PPPoE so you lose about 10%
>with protocol overhead. My 3Meg connection costs
>$50 CDN per month.
Damn! I'm spending too much for my cable ~$45 US and I get 1.5M/128k
(I can't verify the 128K).
Thanks for the info on PPPoE, I was going to look at that.
>Business connections you can get 6Mbit down / 1Mbit up,
This is VDSL and is only avaialble at short distances (like 1000 ft or
less, but don't quote me on that).
Neil Cherry
>n...@bgp558256bgs.ewndsr01.nj.comcast.net (Neil Cherry) wrote:
>
>>I think a T2 is ~6M (96 channels), I think the only people to see
>>those are CO techs and maybe a few of the lab folks.
>>
>>Compare these numbers to DSL with ~384K/128K (is that correct?) and
>>cable with ~1.5M/128K (at least mine is).
>
>My ADSL is 768 Kbps download / 384 Kbps upload. There are some faster
>offerings. I believe my son's company has about twice my speed. VDSL can be
>more than 10 times the speed of ADSL.
I've been having troubles with keeping DSL speeds in my head, world
wide the bandwidth differs and I mostly work with symmetric speeds.
>But, the telcos are not likely to be eager to replace existing hardware
>since few face any real competition for broadband.
This is correct for the US but in the rest of the world ADSL is
selling like mad.
>And your cable speed deteriorates with the number of users whereas ADSL is a
>one-to-one thing.
While this is true for poorly designed neighborhood connections
(residential) this becomes some what of a fallacy as the problem moves
further into the network for DSL. If the DSL provider doesn't have the
correct bandwidth at their POPs then they'll have exactly the same
problem, one user hogging up the bandwidth. Cable providers are now
pushing all their modems to be DOCSIS 1.1 or better compliant so they
can do bandwidth control at the modem. Most have fixed the config
problem that allowed the user to modify the configs. So over all the
cable/xDSL arguement is now almost non-existant.
I'll have more info on xDSL over the next few months, first I have to
program my DSLAM to build up my in lab xDSL network. Maybe I'll be
able to convince work to allow me to get an SDSL connection to my home
as part of my lab experiments, hmm. ;-) Too bad it won't be able to
get to the internet. :-(
J. Clarke
cha87...@projectplasma.com says...
> On Tue, 24 Sep 2002, J. Clarke wrote:
>
> > And enterprise class fibre channel drives don't ever fail?
>
> "Enterprise" class hardware has quite a bit of redundancy,
So where is the redundancy in a Seagate Cheetah, which last I heard was
an "Enteprise-class" drive?
> something I want
> for each family member's computer, but adding a RAID to each machine
> is a bit expensive.
A mirror is quite sufficient, and costs the price of a second drive
(about 68 bucks) with Linux or Netware--with Windows you have to buy a
piece of hardware solely due to a Microsoft marketing decision.
However, to refresh your memory, you commented on the lack of
reliability of IDE drives, as if there were some other type of drive
that did not suffer this lack. Do you believe that there is a type of
drive which is completely reliable?
> > You have to be a newbie.
>
> Your dazzling use of hominem-fu has blinded me. I cede the argument.
So you admit you're a newbie? Then you shouldn't pontificate.
1
You're correct. I knew the Connector part didn't sound quite right when
I wrote it.