Best practice on what to do when redisReply object is Null or has unexpected type.

[mmaatuq]

Hi,
As per documentation

The return value of redisCommand holds a reply when the command was successfully executed. When an error occurs, the return value is NULL and the err field in the context will be set (see section on Errors). Once an error is returned the context cannot be reused and you should set up a new connection.

It means when there is an error I should free the redisContext and initiate a new connection. is this the correct way to do it?
I looked at the provided examples "example-push.c"
and found

    reply = redisCommand(c, "GET key:%d", j);
    assertReplyAndFree(c, reply, REDIS_REPLY_STRING);

Inside assertReplyAndFree it checks if a reply is NULL or its type is different and if it's so they call exit(-1)
what if I don't want to exit at this point, in my application I expect different types to happen a lot.
so in this case instead of exiting I'm freeing the rediscontext object and creating a new connection
Am I doing the right thing here, I think this might be too slow to initiate a new connection every time a key is not found in Redis DB.

[mmaatuq]

quick update:
I think we should free and reconnect if the reply is NULL, but if it has a different type we don't have to free and reconnect.