[Tarantool-patches] [tarantool-patches] Re: [PATCH v4 3/4] box: do not evaluate ck constraints on recovery

[Konstantin Osipov]

* Nikita Pettik <korablev at tarantool.org> [19/10/17 17:06]:
> You made is_force_recovery be global. However, there's a lot of functions
> taking force_recovery param. Please, refactor them: remove that argument
> and use instead global status. What is more, you don't set is_force_recovery
> to false after recovery process is finished. What is more, AFAIR we discussed
> replacement of memtx_recovery_state with one global state shared among
> all engines (instead, you added is_recovery_complete).

Why do you think that now that there is a global state it would be
good to use it everywhere?

Besides, given that now there is is_force_recovery and
is_recovery_complete, maybe have:

engine_is_force_recovery() 
engine_is_recovery_complete() 

declared in engine.h, rather than directly access global
variables?


-- 
Konstantin Osipov, Moscow, Russia