Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 25 Jan 2004 10:06:31 +0000 (UTC)
From:      "Bjoern A. Zeeb" <bzeeb-lists@lists.zabbadoz.net>
To:        pi@LF.net
Cc:        freebsd-isp@freebsd.org
Subject:   Re: BGP4 using FreeBSD
Message-ID:  <Pine.BSF.4.53.0401250935340.25498@e0-0.zab2.int.zabbadoz.net>
In-Reply-To: <20040125092813.GE987@complx.LF.net>
References:  <4012E087.4080504@mr0vka.eu.org> <4012E2F2.2000108@keystreams.com> <20040124215019.GD987@complx.LF.net> <40132999.3020906@keystreams.com> <20040125092813.GE987@complx.LF.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 25 Jan 2004, Kurt Jaeger wrote:

Hi,

> > >Disks were never the relevant topic. Basically, those systems
> > >just worked. Yes, they need a little hand-holding, but not because
> > >of the disks.
>
> > Well that may be, but why risk having a hard drive go out? Flash memory
> > most definatley outlasts traditional hard drives.
>
> See the discussioon nanog about this, a few days ago.
>
> http://www.merit.edu/mail.archives/nanog/
>
> The problem with flash is that has its part of wear and tear as well.
> It's not easy to do a real read-only setup. If you don't, some
> part of the flash do will wear out. Flash mem do has a finite number
> of possible writes.

while this is true loading an md_image from a flash disk is quite easy
these days with FreeBSD as is booting such from tftp.

Then your at about the same problematic cisco p.ex. has writing config
back to nvram or flash. But how often would you do this on a cisco and
how many writes can you do on a flash card these days ?

additionaly you can always save the config to a remote system via
tftp/ftp/scp/sftp/cvs/<insert ..> and if you want you can also fetch
them at boot time again (not the thing one really wants but similar
to some boot tftp modulo boot loader and kernel loading; unsure what
linuxbios et. al. could help here).

logging that needs to be saved should go to a remote system anyway.
If you want local accounting (p.ex. with a hacked up tcpdump ;-) you
may want to have a hd for saving the dumps (perhaps hot swap supported
so they can be changed on failure with less/no downtime of your peerings)
or have an extra 512MB RAM for this with the possibility of losing some
accounting on reboot (which should not happen too often for a router).


Just to note there is also a bgpd.pl (BGP daemon implemented in perl) at
	http://sourceforge.net/projects/bgpd
I once had a quick look at it after its initial announcement some
(2-3) years agao. Current state is unknown. Never used it.

-- 
Greetings

Bjoern A. Zeeb				bzeeb at Zabbadoz dot NeT
56 69 73 69 74				http://www.zabbadoz.net/



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.53.0401250935340.25498>