[c-nsp] ASR-1001 bgp memory usage

Gert Doering gert at greenie.muc.de
Thu Sep 22 02:42:20 EDT 2011


Hi,

On Wed, Sep 21, 2011 at 09:37:14PM -0700, Mack McBride wrote:
> You do get multiple processes.  The IOS CLI portion is fairly monolithic but even on the old 6500s you
> can restart the BGP process (you have to remove all configured peers to do it though :( )

What you still can't do (and this is where the journey really needs to
end), is "oh, there's a security problem in the BGP code -> install patch,
restart bgpd, no reboot needed".  

Combined with NSF, that would be a major step forward regarding serious
high availability.

> And don't disparage the redundant processes, it isn't a bad design.

I really can't see why "have two fully monolithic IOS processes run at
the same time" is a serious improvement.  Well, yes, it might be faster
than rebooting to do IOS upgrades - can that be done?  That is, install
new IOS, run the second IOS process with the new code, and then cut over?

> Of course full modularity is something of a myth.
> If the underlying OS has issues then all of the processes are going to have issues.

Oh, please.  When was the last time you have seen such "underlying OS issues"
in Linux or any of the BSDs?  While theoretically true, this is more of a
red herring than a serious issue.

gert
-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert at greenie.muc.de
fax: +49-89-35655025                        gert at net.informatik.tu-muenchen.de
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 305 bytes
Desc: not available
URL: <https://puck.nether.net/pipermail/cisco-nsp/attachments/20110922/f34001fe/attachment.pgp>


More information about the cisco-nsp mailing list