[leo_manager] error occasionally happens on startup and fail to initialize mnesia tables

[mocchira]

I happened to face the log below today,

[E]     [email protected]     2018-01-16 16:23:34.222608 +0900        1516087414      leo_manager_sup:create_mnesia_tables_1/2        358     [{cause,{"Cannot install fallback",{noproc,{gen_server,call,[kernel
_safe_sup,{start_child,{disk_log_sup,{disk_log_sup,start_link,[]},permanent,1000,supervisor,[disk_log_sup]}},infinity]}}}}]
[E]     [email protected]     2018-01-16 16:23:34.224877 +0900        1516087414      null:null       0       Error in process <0.859.0> on node '[email protected]' with exit value:
{{nocatch,{error,"Unexpected msg fallback_receiver_loop",{'EXIT',<0.856.0>,shutdown}}},[{mnesia_bup,fallback_receiver_loop,4,[{file,"mnesia_bup.erl"},{line,705}]},{mnesia_bup,fallback_receiver,2,[{file,"mnesia_b
up.erl"},{line,618}]}]}

The first line including "Cannot install fallback" occasionally appears and usually it doesn't matter for subsequent startup processes but at this time things got different. As shown on the second line, something went wrong in mnesia_bup module and the error line dumped and as a result, mnesia initialization between master/slave failed.

We may have to look into "Cannot install fallback".