gillh

8 ms·

31.
by gillh 3y ago
You should check out the Aperture flow control system - https://github.com/fluxninja/aperture We built a weighted fair queueing system in the Aperture flow control system to help alleviate the API load pressure - http
32.
by gillh 3y ago
Very interesting blog post! Our team has been working intensively in this area for the last couple of years - flow control, load shedding, controllability (PID control), and so on. We have open-sourced our work at - https://githu
33.
by gillh 3y ago
Very interesting blog post! Our team has been working intensively in this area for the last couple of years - flow control, load shedding, controllability (PID control), and so on. We have open-sourced our work at - https://githu
34.
CodeRabbit: AI-Powered Code Reviews(coderabbit.ai)
2 points by gillh 3y ago|0 comments
35.
by gillh 3y ago
AI helps as well. For instance, our company has extensively used coderabbit.ai since its inception.
36.
Traffic Jams in the Cloud: Are Overloads Sabotaging Application’s Reliability?(blog.fluxninja.com)
4 points by gillh 3y ago|0 comments
37.
Dueling AIs: AI reviewing AI generated code
3 points by gillh 3y ago|0 comments
38.
Dueling AIs: AI reviewing AI generated code
5 points by gillh 3y ago|0 comments
39.
Dueling AIs: AI reviewing AI generated code(github.com)
4 points by gillh 3y ago|1 comments
40.
by gillh 3y ago
Came across this very interesting GitHub pull request of an AI bot reviewing another AI bot’s generated code! Is this the way of the future?
41.
by gillh 3y ago
Aperture project's policy language is indeed inspired by Control Systems, especially PID Control. See: https://docs.fluxninja.com/concepts/advanced/circuit
42.
by gillh 3y ago
Thanks for the feedback - This blog could have been better and less aggressively distributed. At the same time, we should consider the credible body of work behind this. It's not AI-generated spam but backed by actual code that provide
43.
by gillh 3y ago
It's a blog on an open-source project that precisely tells you how to implement adaptive rate limiting. Just click around a bit: - https://github.com/fluxninja/aperture - https://docs.fluxninja.com&#x2F
44.
by gillh 3y ago
While the tech these companies built (stateful reconstruction) is quite hard to engineer, large-scale observability platforms like Datadog were much more successful as they could cover a much larger surface area with minimal incremental eff
45.
by gillh 3y ago
Companies that pioneered the stateful reconstruction of APIs from raw network events - - Netsil [0] (acquired by Nutanix) - Academic spin-out from UPenn - Pixie Labs [1] (acquired by New Relic) [0]: https://thenewstack.io/n
46.
by gillh 3y ago
Good observation. We were able to prevent issues due to limited context by sending the entire file in the prompt and the results were pretty amazing. We eventually reverted the change due to 2 reasons - 1. Limited tokens: 8K tokens is somet
47.
by gillh 3y ago
It’s all about providing relevant context in the prompts and using a better model where reasoning is needed, eg gpt-4. If you are curious, some of our prompts are in the open source. Here is the link to our open source version - https:&#x2
48.
Show HN: Review GitHub PRs with AI/LLMs(coderabbit.ai)
2 points by gillh 3y ago|4 comments
49.
Protecting PostgreSQL with Adaptive Rate Limiting(blog.fluxninja.com)
4 points by gillh 3y ago|0 comments
50.
Show HN: AI-driven GitHub PR reviewer(coderabbit.ai)
4 points by gillh 3y ago|0 comments
51.
by gillh 3y ago
Look at all these super knowledgeable “engineers” down voting this. Load shedding is exactly what would prevent this in the future, scaling up capacity and adding a cache is only going to help buy some time. Wish I could have a dollar for e
52.
by gillh 3y ago
Deploy Aperture for runtime protection. https://github.com/fluxninja/aperture
53.
by gillh 3y ago
The need an adaptive protection system like Aperture[0] to mitigate overloads. [0]: https://github.com/fluxninja/aperture
54.
Prioritized Load Shedding (SRECon'23 APAC Slides) [pdf](usenix.org)
2 points by gillh 3y ago|0 comments
55.
Automate PR Reviews with ChatGPT(github.com)
2 points by gillh 4y ago|0 comments
56.
by gillh 4y ago
It’s customized to our policy spec. But you can learn from this and adapt it to your spec. https://github.com/fluxninja/aperture/blob/main/scripts/json...
57.
by gillh 4y ago
We use Jsonnet extensively in Aperture project for generating control policies. - Policy spec is expressed in protobuf format - GRPC gateway plugin is used to generate Swagger (OpenAPI v2) spec from proto files - Jsonnet bindings are genera
58.
by gillh 4y ago
Multi-prompt[0] approach taken while summarizing large number of files and diffs. Prompts has been tuned for a concise response. To prevent excessive notifications, this action can be configured to skip adding review comments when the chang
59.
Show HN: OpenAI GPT based PR reviewer and summarizer(github.com)
1 points by gillh 4y ago|1 comments
60.
by gillh 4y ago
This is probably what HN operators should use! They keep crashing due to overloads and logged-in users.

More ›