Resolved the startup problem by identifying which pg_clog file it was failingon with:
strace postgres --single -D 9.0/data
Then grabbed that file from the replication slave.
Cheersh