I'm having trouble restoring records from an on-prem MongoDB into Azure CosmosDB using mongorestore. It seems that the operation hits Azure's burst limits and leads to records being dropped. I've found that the common suggestion is to provision a much higher number of Request Units (RUs), but I'm not sure if that will solve the problem. I'm looking for any workarounds or solutions that might help me achieve a clean restore without losing records.
2 Answers
Are you utilizing partitions in your setup? When you increase your RUs, it can lead to more partitions. This cycle just continues, as more partitions require even more RUs to manage, which creates even more partitions. It's a bit of a never-ending loop! 😢
Honestly, Cosmos DB can be super pricey, especially when you scale up. It seems great for lower data loads at first, but once you dig deeper, you might find it costs way more than expected.
I just want to get a clean restore! Right now, it feels impossible. This has turned into the most frustrating experience.