HNHacker News
TopNewBestAskShowJobs

xerxes901

321 karma · joined August 27, 2017

submissionscomments
xerxes901··on Hunting down a C memory leak in a Go program
Believe me, before reaching for this stuff there was several passes of staring at the code trying to find something by “thinking it through” as you say. But at the end of the day, we’re human, and fixing bugs by inspection is just not always a realistic strategy.
xerxes901··on Hunting down a C memory leak in a Go program
Off the top of my head I don't think this would work, because Valgrind needs the atexit hooks to run?

Probably some other signal that has a default action of terminate the program that our app isn't handling might've worked though.

xerxes901··on Hunting down a C memory leak in a Go program
Oh hey that looks really interesting. I'll certainly give it a look!

Lesson learned from the Sarama -> librdkafka migration: We actually wrapped Sarama in a home-grown abstraction (to handle some Zendesk-specific concerns). When we made v2 of this wrapper to support confluent-kafka-go, we made sure that no confluent-kafka-go implementation details/types leaked out of the interface. So, in theory, we can change underlying libraries again "easily" (well, with less rework in the apps at least).

xerxes901··on Hunting down a C memory leak in a Go program
We had previously made extensive use of Sarama, another pure-go implementation of the Kafka protocol. After the third or fourth time where I had to use Wireshark to debug a protocol-level bug we figured it would probably be better to use the librdkafka-based library to have better protocol support.

Plus the team running our brokers found that the Sarama clients were a drag on them being able to enable broker features like zstd compression (although I understand Sarama has zstd now). This felt like it was going to be a kind of problem that would keep cropping up.

xerxes901··on Hunting down a C memory leak in a Go program
So, in the interests of full transparency - we at Zendesk are actually running a fork of confluent-kafka-go, which I forked to add, amongst other things, context support: https://github.com/confluentinc/confluent-kafka-go/pull/626

This bug actually happened because I mis-merged upstream into our fork and missed an important call to rd_kafka_poll_set_consumer: https://github.com/zendesk/confluent-kafka-go/commit/6e2d889...

xerxes901··on Call of Duty Adds Kernel Level Driver for Anti-Cheat
I think the idea would be to use the server side cheat detectio to decide which players need to be forced to install the client side cheat detection, basically.

You're right of course, if there's no $$$ to be made from people who won't buy the game if it has client-side anticheat, then why would they develop a whole second system just for people like me who don't want to run this crap, but want to play the game enough that they'll run it anyway?

xerxes901··on Call of Duty Adds Kernel Level Driver for Anti-Cheat
This was a few years ago at least, not sure it was an option back then? Or maybe the idea of futzing around trying to get the binary nvidia drivers working properly on Linux in the middle of a LAN party didn't appeal to me :D

I do worry that this push to ever-more-intrusive forms of anti-cheat will mean that games are less likely to run on Linux. If the source to your anti-cheat kernel module needs to be released under the GPL it's probably not going to be that effective for very long..

xerxes901··on Call of Duty Adds Kernel Level Driver for Anti-Cheat
Because the only people in this lobby will be cheaters lol. Most people don't care about running garbage-tier CoD kernel code on their machines... so I wind up having to because they do, because it's the default.

Although I concede that basically what I'm asking for here is for the default to be _my_ usecase...

xerxes901··on Call of Duty Adds Kernel Level Driver for Anti-Cheat
I think my suggestion handles this situation though. If the cheaters are playing well enough that you can identify them as cheaters, then they would be getting flagged by my hypothetical server-side anticheat as "good" and so need to run the client-side software too.

Also I don't think "twitch streams" are a great sample of "normal level gameplay", most people there would be ranked comparatively highly in MMR I would have thought? That or I'm leaving tons of money on the table by not streaming my wood-tier CS:GO :D

xerxes901··on Call of Duty Adds Kernel Level Driver for Anti-Cheat
The whole "super intrusive anti-cheat" thing is really annoying to me as a non-competitive player, who just plays games like CoD, CS, etc occasionally with friends. And not just academically annoying, but practically too. I used to run Windows virtualised under Linux with a PCIe passthrough GPU. I remember having to re-install windows directly on my machine at a LAN party once because CS:GO couldn't be run in a VM anymore :(

I think there's a solution that would let wood-tier players like myself avoid this kind of intrusive software, whilst keeping higher-level play clean though.

- Implement server-side anticheat detection, that will ban things that look like bots. As other commenters have noted, bots are indistinguishable from good humans from the outside. So, the server-side detection would ban both bots and good humans

- Have _opt-in_ client side anticheat detection. If you have this enabled, then you're exempted from the server-side detection. So, you enable this if you're actually good.

Could even make this part of the ranking mechanics. Maybe you _can't_ enable the client-side anticheat until the server-side one thinks you're sus. Then you get prompted to turn on the client-side support and get a cool badge or something.

xerxes901··on Podman: A Daemonless Container Engine
I did some hacking on crun this week to fix a bug, so maybe I can offer a perspective about why it’s written in C.

A huge part of a container runtime is pretty much just issuing Linux syscalls - and often the new ones too, that don’t have any glibc bindings yet and need to be called by number. Plus, very careful control of threading is needed, and it’s not just a case of calling runtime.LockOSThread() in go - some of these namespace related syscalls can only be called if there is a single thread in the process.

It’s _possible_ to do this in go, of course; runc exists after all. But it’s certainly very convenient to be able to flip back and forth between kernel source and the crun code base, to use exactly the actual structs that are defined in the kernel headers, and to generally be using the lingua franca of interfacing with the Linux kernel. It lowers cognitive overhead (eg what is the runtime going to do here?) and lets you focus on making exactly the calls to the kernel that you want.

xerxes901··on Google Cloud Is Down
Spare a thought for the pleasant Australian early Monday mornings too! Always a rude awakening...
xerxes901··on Google Cloud Is Down
Couldn't load the support console to "me too" this one either!
xerxes901··on GCP PubSub & CloudSQL are Down
Our application is calling GetTopic on a whole bunch of topics at startup to ensure that they were already created, so even though the publish error rate was only 1%, the app couldn't start because it kept failing the GetTopic calls :/
xerxes901··on Show HN: Amqphosting, Managed RabbitMQ service
We run our RabbitMQ cluster with pause_minority as the partition handling strategy. This should eliminate most message loss on partition, no?
← PreviousPage 2 of 2