High Availability HDR suggestions

Version: IDS 10 I was reading about HDR and it looks like in HDR blobs are not replicated to secondary server if they are of type simple large objects. We don't want to move to smart blobs as it requires large code change and we probably don't have enought time to make that change and test it. I know of only one solution for this is to move blobs to tablespace, but is that advisable ? So I have following questions: 1. Is there any other option in which blobs could be replicated using synchronous HDR with blobs in blobspace ? 2. I know it's not a good idea, but I thought I'll write. Is it a good idea to write a trigger to update blobs to the secondary server. I guess not because we are solving for 0 data loss. 3. I think only taplespace and sbspaces can be fragmented. Can blobspace be fragmented ? 4. Very generic question - what's a good HDR strategy ?