[hox]

The biggest issue faced with DynamoDB from my perspective has been the problem with any hosted service - that is, the operational stability is in the hands of others entrusted with operating a massive multi-tenant system, and any outage cannot follow your own operational recovery mechanisms unless you plan for failover yourself. And the moment you plan for failover, you need to evaluate why you don't just handle the primary system as well.

[apetresc]

There are many things people justifiably complain about DynamoDB, but operational support has rarely been one of them.

[hox]

Dynamodb experienced a severe outage in us-east in September of 2015, dropping the service to three nines of uptime on just the single event. During the incident there was little-to-no knowledge of when service would be restored, so companies that relied on the service were flying blind.

Such an outage can sometimes be mitigated when operations are internal to a company; but with DynamoDB, one has to build internal failover or implement a complex cross-region replication scheme. The complexity and reliability starts to reduce the attractiveness of the hosted service in general.

[elvinyung]

I mean... of course? But I feel like the idea here is that you can "entrust" them with operating the system and not worry about rolling your own failover mechanism, as long as they seem to be upholding their SLAs.

(Of course, cloud vs. bare metal is a long-running debate, but I think I take the position that multitenant clouds are generally cheap enough that it's worth the cost for the extra feature velocity, if you're not truly at scale.)