Netflix buffering issues: Boxing fans complain about Jake Paul vs. Mike Tyson

Netflix is good only on streaming ready made content, not live streaming, but;

1. Netflix is a 300B company, this isn't a resources issue.

2. This isn't the first time they have done live streaming at this scale either. They already have prior failure experience, you expect the 2nd time to be better, if not perfect.

3. There were plenty of time between first massive live streaming to second. Meaning plenty of time to learn and iterate.

Yea, the issue here isn't just that they're having issues, it's that they're having the same issues they've had before.

You can't solve your way out of a complex problem that you have created and which wasn't needed in the first place. The entire microservices thing was overly complex with zero benefits

I spoke to multiple Netflix senior technicians about this.

They said that's the whole shtick.

They have the NFL next month on Christmas day. So that'll be a big streaming session but I think it'll be nothing compared to this. Even Twitter was having problems handling the live pirate streams there.

Yeah didn't they crash on love is blind or one of their reality shows recently-ish?

That's a ridiculous statement. PrimeVideo is the leader in terms of sports events streaming over internet and it is composed of hundreds of microservices.

Live streaming is just much harder than streaming, and it takes a years of work and a huge headcount to get something good.

Prime famously undid some amount of their microservices recently because it couldn’t keep up, and was hideously expensive.

It was a single team for a very specific use case.

To be clear when I said that PrimeVideo is composed of hundreds of microservices, I actually meant that it's composed of hundreds of services, themselves composed, more often than not, of multiple microservices.

Depending on your definition of a microservice, my team alone owns dozens.

> Even Twitter was having problems

Is that a surprise? They're not who I would think of first as a gold standard for high viewership live streams.

This comment shows how a very random blog about a very small part of a product can dominate all conversation about it. Prime video famously did not undo anything. Out of 100+ teams one team undid one service. But somehow similar comments are common on HN. I am making no judgement or microservice or not just on this particular comment.

Well, considering it was multiple small streams I would expect them to keep them up. No have their entire streaming service have issues.

The problem is that provisioning vast capacity for peak viewership is expensive and requires long-term commitment. Some providers won't give you more connectivity to their network unless you sign a 12 month deal where you prepay that.

Peak traffic is very expensive to run, because you're building capacity that will be empty/unsused when the event ends. Who'd pay for that? That's why it's tricky and that's why Akamai charges these insane prices for live streaming.

A "public" secret in that network layer is usually not redundant in your datacenter even if it's promised. To have redundant network you'd need to double your investment and it'll seat idle of at 50% max capacity. For 2hr downtime per year when you restart the high-capacity routers it's not cost efficient for most clients.

Then sign a contract with Akamai, who has been in business for 25 years? You outsource if you aren’t planning to do something very often.

There is no middle ground where you commit a mediocre amount of resources, end up with downtime and a mediocre experience, and then go “but we saved money.”

Well, they didn't want to spend the money or more likely their own technical team/boss promised that they can do it themselves.

They indeed have a great CDN network, but it's not very good for this particular type of traffic. May be they will know/fix/buy next time...

When Apple moved off Akamai for their Keynote live streaming, ( I remember they also used Limestone or EdgeCast ) they had some percentage of audience using Akamai and some on their own CDN. I think it took them three years before they completely moved off Akamai. Not sure if that is still case as that was more than 10 years ago.

But like you stated, they dont want to spend money and their technical people couldn't deliver on time. This isn't a technical issue a lot of people on HN and Twitter wants to discuss about. It is a management issue.

Apple was clearly larger than Google when they came out with Apple Maps, and it was issue-laden for a long time. It is not a resource-issue, but a tech development maturity issue.

What's your point? If they couldn't manage to secure the resources necessary, they shouldn't have agreed to livestream it. As a customer, I don't care AT ALL if it's difficult.

As a customer you're right that you don't care. As a company you care for the cost of the product you sell. Companies doesn't care about "the customer" per se, they care about their profit margin and "free market" competition pushes them to lower the price and keep the service level good. Exclusive rights to stream the fight? Well... it better be working, but we won't overprovision and pay over the expected return.

Buying the exclusive rights for 20Mil and puting 30Mil to stream it wouldn't be a very smart choice. Fuckups happen and this might be a mistake that cost them in lost reputation more than they expected to win.

>They already have prior failure experience

What was the previous fail?