If we are a server/accepter, and we get a bad authorizer, discard our old
challenge, since the previous exchange has failed. This way if the client
retries, their new authorize attempt won't conflict with our old challenge.
Signed-off-by: Sage Weil <sage@redhat.com>
return 0;
}
ldout(cct,10) << __func__ << " bad authorizer on " << con << dendl;
+ // discard old challenge
+ auth_meta->authorizer_challenge.reset();
return -EACCES;
}