2 ms·
If I were to make some WAGs about things that could generally be described this way: * increased failure rates of drives (or other hardware) due to thermal cyc
by kube-system 5y ago
If I were to make some WAGs about things that could generally be described this way:
* increased failure rates of drives (or other hardware) due to thermal cycling
* cache power failures leading to corrupted data
* previously unknown failures in recovery processes (like firmware bugs that might be described as a "hardware problem")
* cooling failures leading to hardware failure
Most of these are pretty rare issues, but it becomes more probable to happen at least once when you're running freekin' us-east-1