You're definitely not alone. That indefinite hang without a clear error log is such a classic CloudGen sync headache. A lot of folks here have hit it, and the consensus seems to be you're not missing a setup step - it's an issue with their service's internal state management.
For what it's worth, when we saw configs not propagating, it often came down to their queue order. If you update a pipeline while a sync for it is already queued, it sometimes runs with the old config. We started implementing a manual delay or a "config version check" at the start of our sync jobs as a workaround.
Has your team been able to quantify the failure rate or the manual restart time? That data becomes crucial if you need to push back on the wider rollout.
Automate all the things
Known issue, not a setup problem. The silent hang is their queue system dropping jobs without logging.
Your wider rollout is the real risk. Have you quantified how much engineering time is spent babysitting these jobs instead of actual work? That's the cost they don't put on the spec sheet.
Doubt everything
Yep, the "babysitting tax" is the real cost. We started tagging every Slack intervention in a channel with a simple emoji reaction to log time. It wasn't fancy, but showing that 15-20% of a junior analyst's week was just sync-herding killed the business case for expansion.
>their queue system dropping jobs without logging
We found the logs *were* there, but buried in a separate service audit stream that wasn't linked to the job UI. Had to stitch them together ourselves. Makes you wonder if the obscurity is by design.
Data is the new oil - but it's usually crude.