Vent: My pillar post scheduling system blew up last Tuesday

Had 8 blog posts queued with tags and images, but my automation tool double-posted the same article 4 times during a migration. Had to manually delete duplicates before the analytics got completely messed up. Has anyone else had a scheduling tool just go rogue like that?

3 comments

3 Comments

ben_nguyen1mo ago

Huh, did pausing all the scheduled stuff before the migration actually help? I had a similar nightmare with a plugin that just kept posting drafts it had saved in some hidden cache folder. What finally worked for me was manually flushing the cron jobs on both servers before touching anything else, and then running a SQL query to delete any post status that said 'future' from the old database. Took like 10 minutes but it saved me from having to babysit the whole thing.

the_kevin1mo ago

You mean the old server was still trying to fire off posts while the new one was working? That actually makes sense, but I always thought migrations just moved the data over clean. You'd think the tool would kill the old queue when you disconnect.

gavina731mo ago

Blame the migration, not the tool. Most scheduling apps have a glitch where duplicate queues get created during server handoffs, and if your tool doesn't check for existing posts before publishing, it's gonna mess up. I had a similar thing happen when switching hosts - the old server still had the queue active while the new one started processing it too. Your automation probably saw the same "ready to post" signal from both systems. Next time, pause all scheduled posts before starting a migration, then clear the old server's queue completely before letting the new one take over. Saves you from becoming a manual delete monkey.