[nsp] Re: 12.0(25)S on 7500 platform

[Jared Mauch]

On Thu, May 08, 2003 at 01:59:59PM +0200, Oliver Boehmer (oboehmer) wrote:
> 
> 
> > I had 12.0(25.1)S built for a BGP MED bugfix. I installed it on a
> > core GSR and the result is: it does not send BGP withdraws !!!! I
> > also had it on a VXR and the result is: reboot every 8 hours by BUS
> > error !!! So my recommendation would be: AVOID IT. Assumed of course
> > 25.1S and 25.S were built on the same base-IOS.... 
> 
> 12.0(25.1)S is much different compared to 12.0(25)S as it already
> contains new features, and is totally untested. 12.0S interims releases
> (e.g 12.0(x.y)S) should be used with great care.
> 
> My customer currently runs 25S on GSR in a limited deployment (no
> reported issues), and I know of another large customer testing it
> (successfully so far).

	I've had very bad luck with 12.0(24)S on gsr.  It has an
incredible memory leak that can cause the device to go south in just a
day or two.  We've reverted to [much] older software images to
avoid this issue for now.  It appears it's going to speed our migration
off of this device.

	- Jared

-- 
Jared Mauch  | pgp key available via finger from jared at puck.nether.net
clue++;      | http://puck.nether.net/~jared/  My statements are only mine.