Netflix Hit By Outage On Christmas Eve

[nutzo]

Been saying this for the last few years that my company has been trying to get into cloud computing, and they look at me like I just said I was a Nazi.

Who really thinks it's a good idea to make your data only available via an uncontrollable remote server that is subject to downtime for a litany of reasons?

Exactly. I'd rather depend on my own ability to fix the problem, than have to keep telling management that I don't know know when the "cloud" will be back up.

I use an external Gmail account to text me when there is a problem with the office email, and the last time I was there was a problem, it was due to Gmail being down

 

[BoogerBomb]

It is becoming pretty obvious that the team that founded and grew Netflix needs to be replaced now that Netflix is a larger, more-established company. This is actually quite common in the history of a company. From pricing fiascos, to "Quikster", to the stupendous decision to dependent on a competitor for streaming, the current "leadership" seems determined to kill their own brand.

So Amazon going down is still Netflix's failure. Damn you guys sure do despise Netflix around here.

 

[Aluisious]

Much of the problem with "cloud computing" is that a cloud is an inapt metaphor for distributed computing. What the fuck does a cloud have to do with anything?

 

[MikeTrike]

Much of the problem with "cloud computing" is that a cloud is an inapt metaphor for distributed computing. What the fuck does a cloud have to do with anything?

It sounds catchy to "executive ignoramuses."

 

[beowulf7]

Yet another nail in NF's coffin.

 

[toast0]

I see a lot of remarks blaming cloud computing in general as a bad thing. Its a great thing, IF it meets your needs and you know what you are buying. I have a business that needs significant amounts of elastic computing power. I can afford it by paying for a cloud service what would require a small server farm otherwise. Now, I do have backup reserve retainer with another cloud service for pretty cheap, AND I keep a couple servers (not a server farm) for emergencies.

The biggest problem with cloud computing is the people who don't know what they're buying. A cloud instance is simply not as reliable as a dedicated server in a colo, but a lot of people think they are for some reason. It's a bad idea to run your service from a single colo, but it's a tragically bad idea to run your service from a single cloud availability zone. Netflix is a big enough company that they shouldn't have a single point of failure for their day to day stuff; and we could have a discussion about if Amazon's whole system counts as a single point of failure, but certainly an availability zone does. I'd guess it makes sense (or could make sense) to do transcoding in ec2, because demand for transcoding is probably not very constant, and if there's some cloud issues, they can always do it later and delay content being available, but everything else, I'd be surprised if the demand changes are high enough so that elasticity overcomes the price differential between real hardware.