How does this work with a tool like jujutsu that provides native support for stacking and preferably don't have a separate tool mucking about with state?
one thing to add for passive healthchecking and clientside loadbalancing is that throughput and dilution of signal really matters.
there are obviously plenty of low/sparse call volume services where passive healthchecks would take forever to get signal, or signal is so infrequently collected its meaningless. and even with decent RPS, say 1m RPS distributed between 1000 caller replicas and 1000 callee replicas, that means that any one caller-callee pair is only seeing 1rps. Depending on your noise threshold, a centralized active healthcheck can respond much faster.
There are some ways to improve signal in the latter case using subsetting and aggregating/reporting controllers, but that all comes with added complexity.
From a dataplane perspective, it does mean your healthchecks are running from a different location than your proxy. So there are risks where routability is impacted for proxy -> dest but not for healthchecker -> dest.
For general reliability, you can create partitions of checkers and use quorum across partitions to determine what the health state is for a given dest. This also enables centralized monitoring to detect systemic issues with bad healthcheck configuration changes (i.e. are healthchecks failing because the service is unhealthy or because of a bad healthchecker?)
In industry, I personnaly know AWS has one or two health-check-as-a-service systems that they are using internally for LBs and DNS. Uber runs its own health-check-as-a-service system which it integrates with its managed proxy fleet as well as p2p discovery. IIRC Meta also has a system like this for at least some things? But maybe I'm misremembering.
* for client-side load balancing, it's entirely possible to move active healthchecking into a dedicated service and have its results be vended along with discovery. In fact, more managed server-side load balancers are also moving healthchecking out of band so they can scale the forwarding plane independently of probes.
* for server-side load balancing, it's entirely possible to shard forwarders to avoid SPOFs, typically by creating isolated increments and then using shuffle sharding by caller/callee to minimize overlap between workloads. I think Alibaba's canalmesh whitepaper covers such an approach.
As for scale, I think for almost everybody it's completely overblown to go with a p2p model. I think a reasonable estimate for a centralized proxy fleet is about 1% of infrastructure costs. If you want to save that, you need to have a team that can build/maintain your centralized proxy's capabilities in all the languages/frameworks your company uses, and you likely need to be build the proxy anyways for the long-tail. Whereas you can fund a much smaller team to focus on e2e ownership of your forwarding plane.
Add on top that you need a safe deployment strategy for updating the critical logic in all of these combinations, and continuous deployment to ensure your fixes roll out to the fleet in a timely fashion. This is itself a hard scaling problem.
They invented a language to avoid you imperatively updating infrastructure, but that's not what CDKTF does; it just makes it easier to materialize that declarative output.
It also makes it easier to reason about that output as you can avoid awkward iteration in your declarative spec.
my expectation is that they would either sell crucial RAM at such a low volume and/or such a high price that it would do more damage to the brand than sunsetting it and returning to it when the slowdown occurs.
Yes. the acknowledgement you're getting in your application code is that the data was persisted in sqlite on that host. There's no mechanism to delay acknowledgement until the write has been asynchronously persisted elsewhere.
litestream makes very few consistency guarantees compared to other datastores, and so I would expect most any issues found would be "working as intended".
at the end of the day with litestream, when you respond back to a client with a successful write you are only guaranteeing a replication factor of 1.
> the last place couldn’t because datadog apparently bills sidecar containers as additional hosts so using sidecar proxy would have doubled our datadog bill.
the problem is that they want to apply a number of stateful/lookaside load balancing strategies, which become more difficult to do in a fully decentralized system. it’s generally easier to asynchronously aggregate information and either decide routing updates centrally or redistribute that aggregate to inform local decisions.
My cheeky answer to "how should this be regulated?" is that sports betting isn't materially different from other high-risk private investments, so it should only be available to accredited investors. Imagine if fanduels/draftkings had to verify assets and income before taking a single bet?!
"Estimates of the scope of illegal sports betting in the United States range anywhere from $80 billion to $380 billion annually, making sports betting the most widespread and popular form of gambling in America."
"AGA’s report estimates that Americans wager $63.8 billion with illegal bookies and offshore sites at a cost of $3.8 billion in gaming revenue and $700 million in state taxes. With Americans projected to place $100 billion in legal sports bets this year, these findings imply that illegal sportsbook operators are capturing nearly 40 percent of the U.S. sports betting market."
I think what would be more interesting to me is estimates on the unique number of citizens betting. Is it up? If so, how appreciably?
How much of this was how bazel works vs. starlark itself?
I find the starlark language is very simple (though inconsistent between the various implementations in bazel, go, and rust) but it takes a bit to understand how the magic between defining rules and implementations works in bazel. and TBH, that is also one place I've really needed auto-completion/static typing in starlark to understand what I can/cannot do.
> I do disagree that there are opportunities to benefit from distributed consensus
there's some benefits to static stability and grey failure, but sure, whatever. the important bit is to have clear paths of aggregation and dissemination in your system.
that being said
> it doesn't matter what some server in SJC says it's hosting
it kind of does matter doesn't it? assuming that server in SJC is your forwarding proxy that does your global loadbalancing, what that server is aware of is highly relevant to what global actions you can take safely.
this doesn't paint a full picture of your options, as there's nothing that stops you from having zonal/regional consensus and then replication across regions/long-range topologies for global distribution.
to be pithy about it, going full-bore gossip protocol is like going full-bore blockchain: solves a problem, introduces a lot of much more painful problems, and would've been solved much more neatly with a little bit of centralization.
Yes, but jquery is much smaller than an empty scala.js app. jquery is about 25kb gzipped and minified (the other comment used the raw size) whereas scala js is 45kb for an empty application.