[bananapub]

in case people don't read all the way to the end, the important takeaway is "you simply can't afford to do instant global state distribution" - you can formal method and Rust and test and watchdog yourself as much as you want, but you simply have to stop doing that or the unknown unknowns will just keep taking you down.

[tptacek]

I mean, the thing we're saying is that instant global state with database-style consensus is unworkable. Instant state distribution though is kind of just... necessary? for a platform like ours. You bring up an app in Europe, proxies in Asia need to know about it to route to it. So you say, "ok, well, they can wait a minute to learn about the app, not the end of the world". Now: that same European instance goes down. Proxies in Asia need to know about that, right away, and this time you can't afford to wait.

[vlovich123]

> Now: that same European instance goes down. Proxies in Asia need to know about that, right away, and this time you can't afford to wait.

But they have to. Physically no solution will be instantaneous because that’s not how the speed of light nor relativity works - even two events next to each other cannot find out about each other instantaneously. So then the question is “how long can I wait for this information”. And that’s the part that I feel isn’t answered - eg if the app dies, the TCP connections die and in theory that information travels as quickly as anything else you send. It’s not reliably detectable but conceivably you could have an eBPF program monitoring death and notifying the proxies. Thats the part that’s really not explained in the article which is why you need to maintain an eventually consistent view of the connectivity. I get maybe why that could be useful but noticing app connectivity death seems wrong considering I believe you’re more tracking machine and cluster health right? Ie not noticing an app instance goes down but noticing all app instances on a given machine are gone and consensus deciding globally where the new app instance will be as quickly as possible?

[tptacek]

A request routed to a dead instance doesn't fall into a black hole: our proxies reroute it. But that's very slow; to deliver acceptable service quality you need to minimize the number of times that happens. So you can't accept a solution that leaves large windows of time within which every instance that has gone down has a stale entry. Remember: instances coming up and down happens all the time on this platform! It's part of the point.

[bananapub]

apologies for misinterpreting it. that said, I would be very interested if in a couple of years you write a followup post about whether you have found that global ~instantaneous state is workable under the right circumstances or not.

[__turbobrew__]

> Proxies in Asia need to know about that, right away, and this time you can't afford to wait.

Did you ever consider envoy xDS?

There are a lot of really cool things in envoy like outlier detection, circuit breakers, load shedding, etc…

[tptacek]

Nope. Talk a little about how how Envoy's service discovery would scale to millions of apps in a global network? There's no way we found the only possible point in the solution space. Do they do something clever here?

What we (think we) know won't work is a topologically centralized database that uses distributed consensus algorithms to synchronize. Running consensus transcontinentally is very painful, and keep the servers central, so that update proposals are local and the protocol can run quickly, subjects large portions of the network to partition risk. The natural response (what I think a lot of people do, in fact) is just to run multiple consensus clusters, but our UX includes a global namespace for customer workloads.

[__turbobrew__]

I haven’t personally worked on envoy xds, but it is what I have seen several BigCo’s use for routing from the edge to internal applications.

> Running consensus transcontinentally is very painful

You don’t necessarily have to do that, you can keep your quorum nodes (lets assume we are talking about etcd) far enough apart to be in separate failure domains (fires, power loss, natural disasters) but close enough that network latency isn’t unbearably high between the replicas.

I have seen the following scheme work for millions of workloads:

1. Etcd quorum across 3 close, but independent regions

2. On startup, the app registers itself under a prefix that all other app replicas register

3. All clients to that app issue etcd watches for that prefix and almost instantly will be notified when there is a change. This is baked as a plugin within grpc clients.

4. A custom grpc resolver is used to do lookups by service name

[tptacek]

I'm thrilled to have people digging into this, because I think it's a super interesting problem, but: no, keeping quorum nodes close-enough-but-not-too-close doesn't solve our problem, because we support a unified customer namespace that runs from Tokyo to Sydney to São Paulo to Northern Virginia to London to Frankfurt to Johannesburg.

Two other details that are super important here:

This is a public cloud. There is no real correlation between apps/regions and clients. Clients are public Internet users. When you bring an app up, it just needs to work, for completely random browsers on completely random continents. Users can and do move their instances (or, more likely, reallocate instances) between regions with no notice.

The second detail is that no matter what DX compromise you make to scale global consensus up, you still need reliable realtime update of instances going down. Not knowing about a new instance that just came up isn't that big a deal! You just get less optimal routing for the request. Not knowing that an instance went down is a very big deal: you end up routing requests to dead instances.

The deployment strategy you're describing is in fact what we used to do! We had a Consul cluster in North America and ran the global network off it.

[__turbobrew__]

> I'm thrilled to have people digging into this, because I think it's a super interesting problem

Yes, somehow this is a problem all the big companies have, but it seems like there is no standard solution and nobody has open sourced their stuff (except you)!

Taking a step back, and thinking about the AWS outage last week which was caused by a buggy bespoke system built on top of DNS, it seems like we need an IETF standard for service discovery. DNS++ if you will. I have seen lots of (ab)use of DNS for dynamic service discovery and it seems like we need a better solution which is either push based or gossip based to more quickly disseminate service discovery updates.

[__turbobrew__]

> you still need reliable realtime update of instances going down

The way I have seen this implemented is through a cluster of service watcher that ping all services once every X seconds and deregister the service when the pings fail.

Additionally you can use grpc with keepalives which will detect on the client side when a service goes down and automatically remove it from the subset. Grpc also has client side outlier detection so the clients can also automatically remove slow servers from the subset as well. This only works for grpc though, so not generally useful if you are creating a cloud for HTTP servers…

[justinparus]

The solutions across different BigCorp Clouds varies depending on the SLA from their underlying network. Doing this on top the public internet is very different than on redundant subsea fiber with dedicated BigCorp bandwidth!

[otterley]

Lots of solutions appear to work in a steady-state scenario—which, admittedly, is most of the time. The key question is how resilient to failure they are, not just under blackout conditions but brownouts as well.

Many people will read a comment like this and cargo-cult an implementation (“millions of workloads”, you say?!) without knowing how they are going to handle the many different failure modes that can result, or even at what scale the solution will break down. Then, when the inevitable happens, panic and potentially data loss will ensue. Or, the system will eventually reach scaling limits that will require a significant architectural overhaul to solve.

TL;DR: There isn’t a one-size-fits-all solution for most distributed consensus problems, especially ones that require global consistency and fault tolerance, and on top of that have established upper bounds on information propagation latency.

[hedgehog]

Is it actually necessary to run transcontinental consensus? Apps in a given location are not movable so it would seem for a given app it's known which part of the network writes can come from. That would require partitioning the namespace but, given that apps are not movable, does that matter? It feel like there are other areas like docs and tooling that would benefit from relatively higher prioritization.

[tptacek]

Apps in a given location are extremely movable! That's the point of the service!

[hedgehog]

We unfortunately lost our location with not a whole lot of notice and the migration to a new one was not seamless, on top of things like the GitHub actions being out of date (only supporting the deprecated Postgres service, not the new one).