IR8x9 power outage

Testing the latest firmware/GOS combination that was provided to us regarding lxc container installation: 

Firmware: ir800-universalk9-bundle.SSA.156-2.0.49.GB

GOS: ir800-ioxvm-1.1.0.4-T.bin 

Everything seems fine with the container software, but experiencing weird issues with the file system after a power outage. 

After a hard reboot (powering off and on) of the router, the container appears as deployed but not activated, and the system does not recover. 

See below the errors while trying to actívate: 

moreno@moreno-VirtualBox:~/IR829$ ./ioxclient app list

Currently using profile :  IR829

Command Name: application-list

List of installed apps :

1. sc         --->   DEPLOYED

moreno@moreno-VirtualBox:~/IR829$ ./ioxclient app console sc

Currently using profile :  IR829

Command Name: application-console

Error. Server returned 500

{

"description": "Cannot console while in DEPLOYED state. Allowed operations are ['undeploy', 'upgrade', 'download_data', 'activate']",

"errorcode": -1019,

"message": "Error getting application console information"

}

moreno@moreno-VirtualBox:~/IR829$ ./ioxclient app activate sc payload device_mapping.json

Currently using profile :  IR829

Command Name: application-activate

Could not complete your command :  Error. Server returned 500

{

"description": "Error in mounting app ext2 image: Error cleaning up disk(/software/caf/work/repo/sc/extract_archive/app.ext2): /software/caf/work/repo/sc/extract_archive/app.ext2 is mounted.\ne2fsck: Cannot continue, aborting.\n\n\n",

"errorcode": -1030,

"message": "Error while changing app state"

} 

So unable to actívate the container (it should activate automatically). The only way of recovering of this situation is uninstalling, rebooting the router and installing it again, but this is not an alternative for us in a production system.

It seems after the reboot the ext2 file system cannot be mounted.

What can we do in this situation. May it be related to our software and we can address it somehow, or is directly an issue with power-cycling the router and the file system.

Have hard booted the devices dozens of times haven't seen this issue before, but lets dig in. 

First, try verifying your images on the device with with: 

# verify /md5 flash:ir800-universalk9-bundle.SSA.156-2.0.49.GB {MD5_HASH_HERE}

# verify /md5 flash:ir800-ioxvm-1.1.0.4-T.bin {MD5_HASH_HERE} 

Against the MD5s provided on the software.cisco.com download page for the images or from whom you got the image.  If you got your images from software.cisco.com, you can highlight over the images on the download page for the details and the MD5 checksum.  Here is an example for the Fog Node image: 

# verify /md5 flash:ir800-ioxvm-1.1.0.4-T.bin b0185c31e33dec17fc46e966836bf636 

Additionally, when you say uninstall, did you mean uninstalling the GOS image?  As in the following: 

# guest-os 1 image uninstall 

And then trying to reinstall the GOS image with: 

# guest-os 1 image install flash:ir800-ioxvm-1.1.0.4-T.bin verify

# guest-os 1 start

Check these images are valid to get started.

• Verified the images and the fog node one md5 hash is fine. What was meant by uninstall was uninstall and reinstall the container itself with the iox tool, not the GOS. The GOS seems to be working, but the logs are showing errors mounting the ext2 filesystem of the container. Therefore, must uninstall and reinstall the container (our app) in order to be able of working again.

You may want to return to the technician from which you have received the core image from and verify the issue with them.  There might be a compatibility or defect in that version of the image.  Again, yet to see the behavior you have reported on the official images from software.cisco.com.