Booting LPWA modem kernel fails after seemingly successful update to 2.0.30

[ivajloip]

Hello,

A few days ago I tried to upgrade the firmware of a Cisco LoRaWAN Gateway from version 2.0.20 to version 2.0.30 following this guide and using serial console. It seemed that everything went fine with the upgrading, however when I rebooted the gateway, it got stuck on what seemed to be the boot loader (see the attached log file for complete reboot logs). The error message looks like

Unsupported hash algorithm for 'hash@1' hash node in 'kernel@1' image node

Bad Data Hash

I tried to put the image on a USB flash drive, but I couldn't figure out a way to make the gateway do anything with it.

Needless to say, we would like to have our gateway operational as soon as possible and any advice in this direction will be greatly appreciated!

Thank you in advance!

[vnallamo]

Upgrading 2.0.30 , needs a uboot upgrade as well. The upgrade procedure used doesn't support uboot upgrade.

Hence there was a mismatch in the version supported and implemented. 

Please use archive download CLI command to upgrade "archive download-sw firmware /normal /save-reload usb: image.xx”  This will take care of the uboot and firmware upgrade together.

Regards

-Vikram

[Naotaka]

Hi.

After trying upgrade to the 2.0.32 , I have a similar situation.

Please tell me how to repair from this situation.

the following log is always appear when reset the LPWA modem.

---

Marvell>> reset
resetting ...

BootROM - 1.73
Booting from SPI flash, Secure mode
BootROM: RSA Public key verification PASSED
BootROM: CSK block signature verification PASSED
BootROM: Boot header signature verification PASSED
BootROM: Box ID verification PASSED
BootROM: JTAG is disabled

General initialization - Version: 1.0.0
AVS selection from EFUSE disabled (Skip reading EFUSE values)
Overriding default AVS value to: 0x23
mvSysEnvIsFlavourReduced: TWSI Read of 'flavor' failed
Detected Device ID 6810
High speed PHY - Version: 2.0

Initialize DB-GP board topology
Device 6810 supports only 2 GbE ports: SGMII-2 @ lane5 disabled (setting USB3.0 H1 instead)
updateTopologySatR: TWSI Read of 'gpserdes1/2' failed
Device 6810 does not supports SerDes Lane #4: replaced topology entry with lane #5
Device 6810/20 supports only 2 SATA interfaces: SATA Port 3 @ lane3 disabled
board SerDes lanes topology details:
| Lane # | Speed| Type |
------------------------------|
| 0 | 5 | PCIe0 |
| 1 | 3 | SATA0 |
| 2 | 3 | SATA1 |
| 5 | 5 | USB3 HOST1 |
-------------------------------
PCIe, Idx 0: detected no link
High speed PHY - Ended Successfully
DDR4 Training Sequence - Ver TIP-0.21.(Sublib 0.5)0
DDR4 Training Sequence - Switching XBAR Window to FastPath Window
DDR Training Sequence - Start scrubbing
DDR Training Sequence - End scrubbing
DDR4 Training Sequence - Ended Successfully
Not detected suspend to RAM indication
BootROM: Image checksum verification PASSED
BootROM: Boot image signature verification PASSED
mvEgigaInit: egiga0 mvNetaPortEnable failed (error)
mvEgigaInit: egiga0 failed
mvEgigaInit: egiga0 mvNetaPortEnable failed (error)
mvEgigaInit: egiga0 failed
mvEgigaInit: egiga0 mvNetaPortEnable failed (error)
mvEgigaInit: egiga0 failed
Booting LPWA modem kernel...
Wrong Image Format for bootm command
ERROR: can't get kernel image!
No previous OS finded!
Marvell>>

[Naotaka]

This problem has solved by myself.

As with earlier indications, U-boot version and Image version were unmatched.

After the old Image was downloaded again , the firmware upgrade was performed again, it succeeded!!

Thanks.

[mrana0101]

Hi

We have the same problem after an attempted upgrade to version 2.0.32. 

As you mentioned, downgrading to previous version would restore. Just wondering which commands we can use to download the previous version from USB and/or TFTP server? 

We are stuck in below prompt and not being able to find any documentation how to use or run the commands in this prompt:

Booting LPWA modem kernel...
Wrong Image Format for bootm command
ERROR: can't get kernel image!
No previous OS finded!
Marvell>>

Please advise. Thanks. 

[Naotaka]

Hi.

Actually, There was no command log left because it was trial and error, but it was recovered by the following method.

(1) Download old Image before Update from Cisco Portal
(2) Set up the TFTP server on the same NW of LoRaGW, and place the (1) Image
(3) Set the address with DHCP on the LoRaGW Uboot command, and enable communication with the TFTP daemon.
# dhcp

(4) Set the TFTP server IP
# setenv serverip 192.168.X.X

(5) Use the following command to boot TFTP
# setenv bootargs $ console $ nandEcc $ mtdparts root = / dev / ram0 rw initrd = 0x8000000,128M ramdisk_size = 128000000; tftpboot 0x3000000 release.itb; bootm 0x3000000

(X) after you recovery with old image, you should upgrade according to the following URL.

https://www.cisco.com/c/en/us/td/docs/routers/interface-module-lorawan/software/release/notes/b_lorawan_rn_2_0_32.html

Regards,

[mrana0101]

Thanks Naotaka. 

However, in my case the below has worked for me.

Procedure:

First please download a valid IXM image:https://software.cisco.com/download/home/286311296/type/286311234/release/2.0.32  

Untar the file and get 'recovery.itb' & 'release.itb' the and copy to your tftp server  

On the IXM uboot shell, perform the following:  

Network configuration:  

Marvell>>setenv ipaddr 192.168.99.99 

Marvell>>setenv serverip 192.168.99.2 (your tftp server ip)  

Marvell>>setenv netmask 255.255.255.0  

Test the connection to the TFTP-server:  

Marvell>>ping 192.168.99.2

Using egiga0 device  

host 192.168.99.2 is alive  

Set the environment variables to boot the release.itb from your TFTP-server:  

Marvell>>setenv bootargs $console $nandEcc $mtdparts root=/dev/ram0 rw initrd=0x8000000,128M ramdisk_size=128000000  

Marvell>>setenv bootcmd_fit 'tftpboot 0x3000000 release.itb;bootm 0x3000000'  

Marvell>>saveenv  (in case error is returned with command Disabled after this command is entered, then proceed with “run bootcmd_fit” as below. But upon reloading the IXM Gateway, it will fell back to Marvell>> prompt)

Boot the image:  

Marvell>>run bootcmd_fit  

Once booted, reinstall the image using the normal procedure from the CLI

In order to prevent the IXM Gateway falling back to Marvell>> prompt, below do the firmware upgrade with /uboot-factory arguement as below:

Gateway>enable  

Gateway#configure terminal  

Gateway(config)#interface FastEthernet 0/1  

Gateway(config-if)#ip address 192.168.99.99 255.255.255.0  

Gateway(config-if)#exit  

Gateway(config)#ip default-gateway 192.168.99.2

Gateway(config)#  

Gateway#archive download-sw firmware /uboot-factory /save-reload tftp://192.168.99.2/ixm_mdm_i_k9-2.0.32.tar.gz

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!  

After the system rebooting, you can check system version by:  

Gateway>enable  

Gateway#  

Gateway#show version  

or  

Gateway#show inventory

This has fixed the issue on permanent basis for me and gateway is able to sustain the reloads and boots up with IXM software each time.

[Chris1350]

Many Thanks this solved my issue here !
The PoE supply was unstable caused by a loose cable shaken by wind that finally broke, when changing the cable the ixm was no more booting.

[fhadj]

Procedure  works for me as well