Every sunday my reindex job runs and when it goes up against the largest table in db (90,000,000 rows) the reindex takes very long.. not sure exactly but the job which is around 20 reindex commands takes around six hours to complete. While this job is running my replication logreader times out and fails causing manual intervention. DB is 120GB TLog is 25GB Should the tlog be smaller. I've always read that Tlog should be sized around 1/4 the size of the db? Any ideas what to do about replication failure? Thanks!