Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 15 Nov 1996 11:33:42 -0800 (PST)
From:      Dror Matalon <dror@dnai.com>
To:        Joe Greco <jgreco@brasil.moneng.mei.com>
Cc:        isp@FreeBSD.org
Subject:   Re: Router Purchase - the bottom line
Message-ID:  <Pine.NEB.3.93.961115112450.12278T-100000@mars.dnai.com>
In-Reply-To: <199611151452.IAA27765@brasil.moneng.mei.com>

next in thread | previous in thread | raw e-mail | index | archive | help

I'm beginning to buy this like of thinking. 
My original thinking was that I've seen disk fail more than anything
on our FreeBsd servers so that I didn't feel comfortable with the
idea of having a box with a disk running as a router. 
On the other hand the idea of one of our server having a disk failure
doesn't cause fear in my heart the way having our Cisco fail does. 
I know that we have enough parts in house to rebuild almost any
server from scratch in a couple of hours. We can get additional
parts any day of the week, almost any time of the day. On the
other hand if the Cisco's power supply or motherboard die we're
in trouble. We do have dual Ethernets and dual T1s still, it's
a bottleneck. 

So, I guess we'll start considering using T1 cards and FreeBSD boxes,
just need to have a spare T1 card.
Now, if someone would only offer a T3 card we'd be really happy. 


On Fri, 15 Nov 1996, Joe Greco wrote:

> > > Look...all of the mentioned product work fine...if they didnt they none
> > > of us would be in business. The original question of cisco 25XX  vs
> > > ET is a simple one......
> > 
> > Am I the only paranoid person on this list or are there other people
> > that would not feel comfortable running FreeBsd as a router unless
> > they were mirroring the disk. FreeBsd are great servers, but if
> > anything fails it's the disk. I'd rather not have this extra
> > risk on my routers.
> 
> I would like to see support for mirroring of the disk, certainly.
> 
> I have seen disks fail.
> 
> I have seen hubs fail.
> 
> I have seen network cards fail.
> 
> I have seen cables fail.
> 
> It does not really matter what you use for a router...  I have seen
> Ciscos fail due to faulty power supplies (Marquette University), I 
> have seen Ciscos crash due to software problems (a local ISP), I have
> seen Ciscos rendered useless by pilot error.  It is most impressive to
> see a $70,000 7XXX class Cisco sitting useless waiting for spare parts.
> 
> On the flip side, a poorly designed PC is unreliable and cranky.
> 
> If you do not pay attention to power protection, adequate cooling (both
> of the machine room and also inside the machine, ball bearing fans, extra
> fans in the case, good CPU fan, cool running CPU, etc), motherboard
> issues (get a GOOD PCIset from a reputable vendor), etc., your router 
> will be crappy and so will your reliability.  You get what you pay for..
> and often you get less.
> 
> With either a Cisco OR a PC, one needs to have a contingency plan for
> WHEN (not if - WHEN) that piece of equipment fails.  My personal feeling
> is that it's cheaper to stock PC spare parts.  I can use them for other
> machines too if other machines fail.
> 
> There is a "Contingency plan" list at that local ISP, if I remember 
> correctly, their contingency plan for failure of their Cisco 7XXX router 
> is to be screwed until replacement parts arrive.  It says something to 
> that effect.
> 
> My "Contingency plan" was recently tested when a hub died and somehow took
> one of my routers with it.  Unfortunately I was not able to get on site
> for several hours, but I was able to cover the loss of my core router by
> having a tech plug a few critical nets into the spare Ethernet ports I
> am so fond of equipping my other routers with...  it was not too hard to
> reconfigure things so that I was mostly live.  When I got on site, six
> hours later, I took a waiting spare machine, plopped a 4-port PCI Ethernet
> card in it, and 30 minutes later my "all new" core router was back on line.
> 
> I was not happy with the failure of my contingency plan - it required
> having someone present on site who could follow instructions over the
> phone.
> 
> I will soon be deploying redundant routers, with OSPF routing, to help
> eliminate "single points of failure".
> 
> So I think the real deal has more to do with proper contingency planning
> than it does with Cisco routers.  If and when the ISP with the Cisco has
> their 7XXX crap out, they are S.O.L.  If and when one of my routers craps
> out, I will not have to worry too much.  And I will have paid less than
> $70,000 for my solution to the problem.
> 
> You are not the only paranoid person.  Paranoid people are the only people
> who are successful playing this game in the long run.  :-)
> 
> ... JG
> 

Dror Matalon                                            Voice: 510 649-6110
Direct Network Access                                   Fax:   510 649-7130
2039 Shattuck Avenue                                    Modem: 510 649-6116
Berkeley, CA 94704                                      Email: dror@dnai.com




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.NEB.3.93.961115112450.12278T-100000>