Symptoms

Config load fails when attempting to load a config that was saved from before 12.1.0 on a system that was already upgraded. This happens only if before the upgrade, there was an ASM logging profile which had both remote logging and local logging enabled on it. In the case of a single logging profile with local-plus-remote ASM enabled on it, upon an upgrade, the logging profile is split into two profiles. One has the '_local' extension added to it. Another attempt to load the config of the pre-upgrade system will fail. This only happens when using 'load sys config' or 'load sys config file', and does not happen when using 'load sys ucs'. Upon failure, the following error is seen on the terminal: 01070710:3: Cannot update_indexes/checkpoint DB object, class:fw_log_profile status:13 - EdbCfgObj.cpp, line 127 Unexpected Error: Loading configuration process failed. And in /var/log/ltm: err mcpd[6618]: 01070710:3: Database error (13), Cannot update_indexes/checkpoint DB object, class:fw_log_profile status:13 - EdbCfgObj.cpp, line 127.

Impact

Config load fails. Upgrade fails.

Conditions

-- Using a configuration that contains a Log Profile with ASM enabled and both Remote Log and Local Log enabled. -- Upgrade to 12.1.2 or later (Use roll-forward upgrade, or instead use clean install and afterwards load the saved config file).

Workaround

Use one of the following Workarounds: 1. Save the new configuration before editing and re-loading, using the following commands: tmsh save sys config partitions all tmsh load sys config partitions all (Note: Saving the UCS also saves the configuration.) 2. Instead of loading the full configuration directly, first load the base and then load the full configuration: tmsh -c 'load sys config partitions all base; load sys config partitions all'

Fix Information

None

Behavior Change