psch

Also, I might suggest you handle the cron task differently. The duplication bug is exaggerated by the way the job is set up. It never completes the update, so the time stamp of next run doesn’t get updated – it ends up scraping on every page request. Static “once every 2 hours” would probably suffice.