[66280] in North American Network Operators' Group
Re: GSR, 7600, Juniper M?, oh my!
daemon@ATHENA.MIT.EDU (Rob Healey)
Wed Jan 7 13:13:36 2004
From: Rob Healey <rhealey@benjy.onvoy.com>
In-Reply-To: <3FFB96C5.20503@utc.edu>
To: Jeff Kell <jeff-kell@utc.edu>
Date: Wed, 7 Jan 2004 12:12:51 -0600 (CST)
Cc: nanog@nanog.org
Reply-To: rhealey@norstar.com
Errors-To: owner-nanog-outgoing@merit.edu
> > Never under any condition let anyone tell you that Juniper is perfect...
> > But, as everyone that uses both will tell you, it is "better" (at most
> > things).
>
> They tend to be (in our experience) a "set it and forget it" thing,
> while you can spend considerable time tweaking your Cisco; but
> admittedly ours is just an M5 edge (two gig-Es into an OC48). The rest
> of our gear is Cisco, and yes, we draw straws to see who has to mess
> with the Juniper when it has to be messed with.
>
If you can set aside time to study JUNOS a bit perhaps the straw draws
won't be necessary! Although you are correct about the "set and forget"
nature of JUNOS vs other platforms, especially when the larger DDOS's
hit...
The 1980's origin of IOS becomes clear when you look at more modern
systems like JUNOS, especially if you've had structured and/or
object oriented programming.
Many interesting network solutions that have to be dismissed outright
because of IOS limitations, weaknesses or bugs can be easily expressed
in newer systems, not just JUNOS.
If you have the time to give them a look, the non-IOS systems have
ALOT to offer in terms of expressing new/innovative ways to solve
networking design/arch problems.
Its too bad Juniper never released the old "Olive" JUNOS's for
general download; they worked on stock x86 hardware with Intel
and 3com(?) ethernet interfaces. They were GREAT for learning
JUNOS on. Maybe Juniper will rethink that decision
for marketing gains of exposing more people to JUNOS?
-Rob
