Strategy to get support for non-user-replicable problems?

I’m hoping we can keep this one constructive…? Support is always a subject prone to derailing into a pile of rants, but it would be really nice to talk about a solution.

I’m slowly garnering a small collection of things that I can’t reproduce on demand for various reasons. They’re actually demonstrably broken things that cause real problems. Sometimes it’s a server process that randomly borks its work 5% of the time, even over identical tasks. Sometimes it’s something that could be demonstrated but there isn’t a safe way to do so. SaaS users are especially limited when it’s an intermittent server side problem that’s completely out of our control to observe or reproduce.

All respect to support - I get the impression that they don’t have an approved workflow for these problems, and that they’re frustrated too. I know from experience that trying to find solutions within an inflexible process framework inevitably evolves into really awful place to be as unsolved problems accrue. Is there a constructive way to help them help us?

3 Likes

I’m going continue to encourage better telemetry. If we can capture logs, traces, and metrics locally or in the cloud, then support will have the required information to hand over to development for those non-reproducible issues. Having this would help users capture the conditions when these events occur as well.