Replacing a Dual Supervisor Module in a 6513 switch

[aalbrecht27]

I have a supervisor blade, (VS-S720-10G-3C), in a 6513 that is faulty and needs replaced. Do I need to replace the IOS on the new supervisor blade to make sure it is the same as the IOS image that is on the Active supervisor module, or will it synchronize automatically?

In the synchronization process, the active supervisor engine checks  the  standby supervisor engine run-time image to make sure that it  matches  its own run-time image. The active supervisor engine checks  three  conditions:

•If it needs to copy its boot image to the standby supervisor engine

•If the standby supervisor engine bootstring needs to be changed

•If the standby supervisor engine needs to be reset

I was reading the above documentation that mentioned the active supervisor should copy the image if it detects a newly installed standby supervisor blade with a different image, is this correct or was it referring to synchronizing the configuration file?

Message was edited by: Aaron Albrecht

[Ivan Shirshin · Accepted solution]

Hi Aaron,

Which OS are you running  - CatOS or IOS? The behavior is different.

CatOS would attempt to copy the images to standby to match and you can see 4 different examples of different outcomes based on the swith setup and flash conditions:

By default on the Cisco Catalyst 6500 Series  Switch, the Catalyst OS software images on the active and the standby  Supervisor Engines must be the same. If the two supervisor images are  not the same version during system bootup, the active Supervisor Engine  downloads its current boot image to the standby Supervisor Engine. The  NVRAM configuration of the active Supervisor Engine is also synchronized  between the Supervisor Engines.

http://www.cisco.com/en/US/products/hw/switches/ps708/products_configuration_example09186a0080a98f3c.shtml#Sup-redund

http://www.cisco.com/en/US/docs/switches/lan/catalyst6500/catos/8.x/configuration/guide/redund.html#wp1107363

But if you are using IOS images on the switch (Native Mode), it won't copy the images and simply change to RPR redundancy mode (less syncronized, longer downtime in case of failure):

The active Supervisor Engine checks the image  version of the redundant Supervisor Engine when the redundant Supervisor  Engine comes online. If the image on the redundant Supervisor Engine  does not match the image on the active Supervisor Engine, RPR mode is  used.

In that case you have to make sure manually that the images match.

Kind Regards,
Ivan

**Please grade this post if you find it useful.

[aalbrecht27]

Sorry, I forgot to mention it is running IOS 12.2(33)SXH5

If we do need to update the IOS where should it be copied to?  This is the part I'm worried about because I want to make sure I copy the IOS to the correct supervisor module.

We haven't yet removed the old supervisor blade that is in standby, but the 'show bootvar' output currently shows:

switch#sh bootvar

BOOT variable = sup-bootdisk:,1;

CONFIG_FILE variable does not exist

BOOTLDR variable does not exist

Configuration register is 0x2102

Standby is up

Standby has 1048576K/8192K bytes of memory.

Standby BOOT variable = sup-bootdisk:,1;

Standby CONFIG_FILE variable does not exist

Standby BOOTLDR variable does not exist

Standby Configuration register is 0x2102

Where would sup-bootdisk:,1: be located on the switch?  Is this on the active or standby module?

I'm a bit confused as to the difference between the directories sup-bootdisk: and slavesup-bootdisk:

They both contain the same IOS version so I'm not sure if one is referencing the active or standby module.

[Ivan Shirshin]

Hi Aaron,

You can see all file systems on the switch by running commands "show file systems". I am not sure where you image is located now but I recommend to match the file systems on the supervisors.

Meaning, if you boot from sup-bootdisk on active, the same image should be present on standby.

Standby sup-bootdisk has a name of slavesup-bootdisk when working in a console of active supervisor (again, check the "show file systems").

Of course, you can also choose disk0 on both or bootflash/bootdisk.

In your current configuration the system is configured to search for the first image file in the sup-bootdisk of the corresponding supervisor.

     BOOT variable = sup-bootdisk:,1;

Note that if you just copy new image there and do not set more specific bootstring with

     sh run | i boot

     conf t

          not boot ...

          boot system ...

          end

and do not delete old image, then it will load the old one (as it would be the first in the list if imaged on sup-bootdisk).

Kind Regards,
Ivan

**Please grade this post if you find it useful.