Should water cool faster if is inserted metal canister with ice inside either mix only with ice? Let's say that we have two canister first bigger (metal canister 2l) with 1l of water at 100C, and second smaller (metal canister 1l) with 1l of ice. And we want to cool down boiled water to 50C. So we will insert ice (only ice not hole canister) into boiled water and after some time get wanted water (50C). 
But what if we insert into boiled canister - small canister (hole canister with ice inside), should boiled water be cooled down faster? (cause smaller canister with metal surface have bigger thermal conductivity than pure water)  
 A: If the cold canister you insert has more surface area than all the ice cubes that you would otherwise insert, then we would presume (for most circumstance) that the canister will perform a cooling action more quickly. 
A variance occurs if the canister is filled with ice cubes.  Then you have two surface-area problems; canister-to-larger-vessel, and ice-cubes-to-canister.  Here, the cold-canister walls are assumed to be perfectly conducting of heat.

If any of this seems off-target, please help by refining the grammar in your post.  I often revise the grammar in posts, but here I am not able because there are ambiguities that cannot be resolved.  
A: Ice absorbs heat from the boiling water by melting. That means, if you put ICE in water, it increases the volume of liquid and their by its mass. if you are considering only the temperature(Not volume) initially and finally, the ice dropped in the canister will be faster, because even after melting of ice, cool water + Dropped ice(more surface area) will continue to absorbs heat until It reaches equilibrium(here 50°C) and thru convection.
Ice inside canister starts to melt creating a layer of water between the canister wall and ice their by lowering conductivity and this cannot be avoided by any means.
So your 1st option is quicker.
