[j-nsp] Host 0 boot from Alternate media on M10 Alarm

[Julien Couturier]

> -----Message d'origine-----
> De : juniper-nsp-bounces at puck.nether.net [mailto:juniper-nsp-
> bounces at puck.nether.net] De la part de Jonas Frey
> Envoyé : dimanche 23 décembre 2007 18:41
> À : David Ball
> Cc : juniper-nsp at puck.nether.net; Mohd Irwan
> Objet : Re: [j-nsp] Host 0 boot from Alternate media on M10 Alarm
> 
> Well, its not uncommon that CF goes bad after time. You will usually
> not
> notice this while normal operation since there are very very few
> read/writes to the CF. When you update junos there are alot of
> read/writes to the CF thus this is where 95% of CF failures are to be
> expected. There is no monitoring within JunOS to check if the CF is
> good
> over time...this is only true for the harddisk where SMART is running
> (but it doesnt help all the time...).
> You can manually check the CF with either the routing-engine
> diagnostics
> (read only test) or try "request system snapshot". The later will also
> write to the CF. We have had CF failures which were not detected by
> just
> read tests but all were clearly visible when doing a snapshot.

Both 'fsck' and 'dd' Shell commands (need 'super-user/root' access to the box) are also helpful to check CF sanity.

Cheers,
Julien.