q3k

39 ms·

571.
by q3k 5y ago
Real World Haskell is unfortunately outdated - it predates the 2014 Applicative Monad Proposal.
572.
by q3k 5y ago
I know it's poking fun at the amount of SCREAMING_SNAKE annotations and checkers and overhead to submit a simple CL... but man oh man do I miss Critique.
573.
by q3k 5y ago
Not GP, but: https://research.google/pubs/pub43438/ https://sre.google/sre-book/production-environment/ are the best public documentation entrypoints that I know of.
574.
by q3k 5y ago
The first issue you've linked concerns quota support for ephemeral storage requests/limits - which is not about the limits themselves, but the ability to set a limit quotas per tenant/namespace. Eg., team A cannot use a total
575.
by q3k 5y ago
> Packing everything together is a "selling point" until you find that a service can fill up ephemeral storage and take down other services, or consume bandwidth without limit. Ephemeral storage has resource requests/limit
576.
by q3k 5y ago
> If you want to see the latter part nicely integrated, come to a SIGAUTH meeting and help me argue for it. Invitation accepted! :) I've been dying to see an ALTS-like [1] thing that works with Kubernetes. I really should be able to
577.
by q3k 5y ago
If you don't depend on clusters being 100% available all the time and design your applications to handle cluster-wide outages (which you need to do at Google scale anyway), then simply doing progressive rollouts across clusters is good
578.
by q3k 5y ago
> Or, maybe I read the papers and book wrong. No, that's exactly how it works. You have clusters spanning a datacenter failure domain (~= an AZ), and everything from prod to dev workloads runs on there, with low priority batch jobs
579.
by q3k 5y ago
Yeah, especially in production bare metal clusters. If you want N+2 redundancy that's at least 5 physical machines for just the control plane (etcd & apiservers), more if you don't want to colocate worker nodes with that... Ev
580.
by q3k 5y ago
> Confidential computing (Intel SGX, ARM TrustZone, AMD SEV-SNP) handle this by encrypting the virtual machine memory so that even having full root on the host does not expose vm compute or memory. Google's current confidential comp
581.
by q3k 5y ago
> Your example is a pretty extreme caricature; It's a caricature? I've only recently learned that this is not something that everyone struggles with daily...
582.
by q3k 5y ago
Right, what probably happened is that an `/accounts/me` endpoint got cached. But what GP seems to be asking about is: “Would having your app always encode a user ID into the endpoint have helped?”.
583.
by q3k 5y ago
Mind you though, I'm not blindly advocating for CDNs. :) In fact, I think that the performance gain is not worth the extra complexity for most sites out there. The recent fastly outage, klarna outage, ... all show that CDNs are no silv
584.
by q3k 5y ago
Most CDNs out there maintain a pool of HTTPS keep-alive'd connections to backing services. This eliminates the roundtrip needed for establishing a TCP connection and the roundtrip (or two) needed for setting up TLS. Another reason a no
585.
by q3k 5y ago
> How would you end up getting other people's data in your app if your app only calls that unique URL? From what I understand what happened with this outage, the CDN would still cache /api/account?id=123, and someone with
586.
by q3k 5y ago
And just to make sure it works if misinterpreted, I have both $firstname@$lastname.pl and firstname@lastname.pl :)
587.
by q3k 5y ago
> Tell that the huge clients that lost sales because of this, and all you have to say is: "wE DoN'T kNoW..." Tell these clients that they should've carefully read their contract with Fastly, especially the 'Servi
588.
by q3k 5y ago
> You've just witnessed almost the entire internet break because of a catastrophic cascading outage that affected lots of huge companies, since third party services used and trusted Fastly. Blame the companies that relied on Fastly
589.
by q3k 5y ago
> Not my problem. Fastly should work as intended. What's your SLA with them? Just assuming things will always work because the marketing copy said so is recipe for disaster. It's hoping that things never go wrong, and when they
590.
by q3k 5y ago
It's empathy towards people managing the incident, not towards the company. It's a sign of solidarity from SRE to SRE, not a sign of solidarity with a company.
591.
by q3k 5y ago
There's, to me, no obvious clear cut line. But here are some indicators that make me consider someone was being grossly negligent and/or even malicious: - ignoring warnings - acting against known-to-them best practices - repeating
592.
by q3k 5y ago
> I disagree. People implemented those systems, so if you are correct that it is the systems fault, then it is also a persons fault. How do you make sure that mistakes don't happen, then? Do you blame and fire people who make mistak
593.
by q3k 5y ago
The point isn't to dance around the incident, but to not blame people. You can blame systems, design, engineering culture, processes, but don't blame people. Even if someone accidentally pressed the 'destroy prod' button
594.
by q3k 5y ago
I also saw a glimpse of 'I/O error'. That sounds fun.
595.
by q3k 5y ago
I also believe this could've been an honest mistake. There's tons of royalty-free texture CD packs from the '90s and '00s, and this could've been confused for one of them.
596.
by q3k 5y ago
You're welcome :). They're also very easy to write, so if you ever happen to run k8s and need to give workloads access to some odd/custom host hardware, implementing a proper plugin for it is quite painless and gives much bet
597.
by q3k 5y ago
Sure, that's one of the cases when this might needed in prod (although in the parent post I meant only access to honest-to-god data files, not things like bindmounting /dev). In practice bindmount smell can also be somewhat allevi
598.
by q3k 5y ago
I would absolutely say it's a production antipattern to run a container with access to some already existing host files belonging to some other user. However, this is something that's basically unavoidable if you're attemptin
599.
by q3k 5y ago
> You can go the extra mile and turn the UID:GID into build args like the original parent and you're good to go. That doesn't help you if you're attempting to use pre-built/existing Docker images that are not built in
600.
by q3k 5y ago
> The only tedious thing is you have to adapt this for every image type you run. The tedious thing is that this escalates into complexity whenever you have to deal with K developers using M projects developed by N teams each using a diff

More ›