Forum Discussion
NimbostratusCannot restore factory defaults after using 11.5
Dears, I have upgraded my F5 appliances to the new software 11.5 which was releases three days ago , after that the active/standby appliances didn't see each other and they both appear as disconnected and the provisioning was disappeared and i cannot see any of the modules.
I'm trying to reset one of them to the factory defaults , however, when i load the default config i get:
Loading configuration... /defaults/defaults.scf Resetting trust domain... Unexpected Error: MCP must be in the running phase to reset the trust domain. Examine the output of
tmsh show sys mcp-stateroot@(localhost)(cfg-sync Standalone)(Offline)(/Common)(tmos) show sys mcp-state
Sys::mcpd State:Running Phase platform Last Configuration Load Status full-config-load-succeed
but now i cannot configure the network section as it gives me the following error: An error has occurred while trying to process your request.
please advise. Ahmad.
17 Replies
Cory_50405Noctilucent
Did you re-license the appliances before upgrading? If not, re-license them now and see if that fixes your issue.
Ahmad_Mohaidat_Yes i reactivated the license , and when i login to the appliance i reach the setup utility , when i try to provision the resources i get the following error : An error has occurred while trying to process your request. btw the device is stuck in "offline" status
- Cory_50405
Have you looked in /var/log/ltm to check for errors to see what the issue might be? Feel free to post any suspicious entries here for the community to comment on.
- boneyard
MVP
it really sounds the device isnt running correctly so then you can perform the load default
do a bigstart status and post the output
nitass_89166is there any error when loading configuration? if yes, i think you may have to fix it before restoring factory default.
tmsh load sys config- Ahmad_Mohaidat_yeah this is what i get: 01070088:3: The requested object name (VeriSign Certificate.crt) is invalid. Unexpected Error: Loading configuration process failed.
- Try to upload cert on your boxes again. You certificate file has space in name, and this may cause the error (files with space in name may not be copied during upgrade). Another workaround you may try is to remove cert name from the config (/config/bigip.conf) manually.
- nitass
Employee
is there any error when loading configuration? if yes, i think you may have to fix it before restoring factory default.
tmsh load sys config- Ahmad_Mohaidat_yeah this is what i get: 01070088:3: The requested object name (VeriSign Certificate.crt) is invalid. Unexpected Error: Loading configuration process failed.
- Try to upload cert on your boxes again. You certificate file has space in name, and this may cause the error (files with space in name may not be copied during upgrade). Another workaround you may try is to remove cert name from the config (/config/bigip.conf) manually.
JGCumulonimbus
Oh, Dear. Why rush for 11.5.0? It's a .0 release anyway.- Ahmad_Mohaidat_we are using a coding language that is not supported on any other software , it was introduced in 11.5.0 and our project was on hold waiting for this software.
- nitass
Try to upload cert on your boxes again.
i do not think gui is workable because configuration cannot be loaded.
Another workaround you may try is to remove cert name from the config (/config/bigip.conf) manually.
i think this is the way.
- Ahmad_Mohaidat_
i worked with F5 Support and the issue resolved after doing a clean install , the bad news is that i lost all my config and i configured the device from the scratch , here are the steps we followed to fix the issue:
- downgrade to 11.4.0
- delete the HD1.3 which has 11.5 installed on in. tmsh delete sys software volume HD1.3
- modify the db key tmsh modify sys db liveinstall.moveconfig value disable tmsh modify sys db liveinstall.saveconfig value disable
- install the 11.5.0 into the HD1.3 tmsh install sys software image <11.5.0 iso file> create-volume volume HD1.3
- reboot to the HD1.5 tmsh reboot volume HD1.3
- JGThanks for sharing the solution. Did Support find what the problem was? Was there an issue ID for this?
- JGI see. So you are going on the bleeding edge. Was it something cloudie? :-)
