Networking: Does slavedisk0 take over for disk0 in the event of a fault?

Question

My bootvar is set to boot up images from disk0. The 6500 series switch has two redundant supervisors. Each supervisor has a PCMCIA card with the exact same IOS images in the 0 slot. The secondary supervisor is given the disk name slavedisk0 and the primary is simply disk0

show bootvar

BOOT variable = disk0:s72033-ipservicesk9_wan-mz.122-18.SXF15a.bin,1;sup-bootflash:s72033-pk9sv-mz.122-17d.SXB11.bin,1;
CONFIG_FILE variable =
BOOTLDR variable =
Configuration register is 0x2102

Standby is up
Standby has 458720K/65536K bytes of memory.

Standby BOOT variable = disk0:s72033-ipservicesk9_wan-mz.122-18.SXF15a.bin,1;sup-bootflash:s72033-pk9sv-mz.122-17d.SXB11.bin,1;
Standby CONFIG_FILE variable =
Standby BOOTLDR variable =

Standby Configuration register is 0x2102

Here is a dir of slavedisk0 and disk0:

Directory of disk0:/

    1  -rw-    80139268  Feb 21 2012 06:30:00 -05:00  s72033-ipservicesk9_wan-mz.122-18.SXF15a.bin
    2  -rw-    42061544  Feb 21 2012 06:35:48 -05:00  s72033-pk9sv-mz.122-17d.SXB11.bin
    3  -rw-       62172  Feb 21 2012 06:37:36 -05:00  file.cfg

256462848 bytes total (134193152 bytes free)
switch#dir slavedisk0:
Directory of slavedisk0:/

    1  -rw-    42061544  Feb 21 2012 05:55:38 -05:00  s72033-pk9sv-mz.122-17d.SXB11.bin
    2  -rw-       62172  Feb 21 2012 05:56:44 -05:00  file.cfg
    3  -rw-    80139268  Feb 21 2012 06:17:06 -05:00  s72033-ipservicesk9_wan-mz.122-18.SXF15a.bin

My question is that in the event the real disk0 fails and the switch reboots, will the router automatically look in slavedisk0? Would my boot statements have to explicitly state to look in the slave and be:

 BOOT variable = disk0:s72033-ipservicesk9_wan-mz.122-18.SXF15a.bin,1;sup-bootflash:s72033-pk9sv-mz.122-17d.SXB11.bin,1;slavedisk0:s72033-ipservicesk9_wan-mz.122-18.SXF15a.bin,1;slavesup-bootflash:s72033-pk9sv-mz.122-17d.SXB11.bin,1;

My only run statements involving boot are:

switch#show run | in boot
boot system flash disk0:s72033-ipservicesk9_wan-mz.122-18.SXF15a.bin
boot system flash sup-bootflash:s72033-pk9sv-mz.122-17d.SXB11.bin

score 1 · Accepted Answer · answered Jan 10 '13 at 05:05

The 'slavedisk0:' location is always on the standby supervisor. You cannot boot off the slavedisk0 file system as the communication between supervisors is not up and running when the switch reboots and the supervisor needs to locate and load code.

If the primary supervisor fails and the other supervisor boots, it sees its own CF disk/internal flash as the 'disk0:' or 'sup-bootflash:' filesystems so you do need to keep the files listed in the boot statement on both supervisors. If the file isn't available on the slave filesystem you will normally see an error message when saving the configuration.

Also, 12.2(18)anything is really old, and you might want to look at upgrading IOS if you have a support contract. If you do upgrade and aren't running WAN modules (OSM/SIP/SPA) the switch will boot faster if you use code without the _wan suffix.

Networking: Does slavedisk0 take over for disk0 in the event of a fault?

1 Answers1