Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
shayonj
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
5 ms
·
1.
▲
by
shayonj
13d ago
Interesting that this is happening around the same time as Claude issues too
2.
▲
by
shayonj
28d ago
Nice one re: flatbuffers in `blob` column. Have been working with flatbuffers a lot and that's a neat idea in general. Not 100% sure about using PG for file system at scale however. I'd love to hear more on the challenges (vacuum,
3.
▲
by
shayonj
1mo ago
v interesting. Thanks for sharing. I kept seeing this phenomenon of "no free lunches". TIL about the waterbed theory.
4.
▲
Every fast write moves work somewhere else
(shayon.dev)
73 points
by
shayonj
1mo ago
|
24 comments
5.
▲
by
shayonj
2mo ago
> Nothing was solved. The four questions got new answers, and the costs moved into the compactor. I have been meaning to write on this topic and in relation to LSM for a while. So, I appreciated that note on LSMs. There are no free ~lunc
6.
▲
Building a Tiny FUSE Filesystem
(shayon.dev)
2 points
by
shayonj
3mo ago
|
0 comments
7.
▲
Exploring building a tiny FUSE filesystem
(shayon.dev)
79 points
by
shayonj
3mo ago
|
15 comments
8.
▲
Building a Tiny FUSE Filesystem
(shayon.dev)
2 points
by
shayonj
3mo ago
|
0 comments
9.
▲
I am not leaving GitHub any time soon
(shayon.dev)
2 points
by
shayonj
5mo ago
|
0 comments
10.
▲
GitHub Down Again
(githubstatus.com)
14 points
by
shayonj
5mo ago
|
2 comments
11.
▲
Shipping Fast Requires a High Degree of Trust
(shayon.dev)
1 points
by
shayonj
5mo ago
|
1 comments
12.
▲
Fast sandboxed code execution with pre-warmed gVisor pools
(github.com)
2 points
by
shayonj
6mo ago
|
0 comments
13.
▲
Fast sandboxed code execution with pre-warmed gVisor pools
(github.com)
2 points
by
shayonj
6mo ago
|
0 comments
14.
▲
by
shayonj
6mo ago
that's very interesting! I was noticing page vault storm on live migrations as well and I wonder if that's what you were running into / mentioning here regarding the lock contention
15.
▲
by
shayonj
6mo ago
> you can do things that aren't practical with 100-200ms startup: speculative parallel execution (fork 10 VMs, try 10 approaches, keep the best), treating code execution like a function call instead of an infrastructure decision, e
16.
▲
by
shayonj
6mo ago
Cool project. +1 on userfaultfd for the multi-node path. Wrote about how uffd-based on-demand restore works wrt to my Cloud Hypervisor change [1] if you are curious. I think the the main things to watch are fault storms at resume (all vCPUs
17.
▲
Linux Page Faults, MMAP, and userfaultfd for faster VM boots
(shayon.dev)
59 points
by
shayonj
6mo ago
|
3 comments
18.
▲
Linux Page Faults, MMAP, and userfaultfd for faster VM restores
(shayon.dev)
2 points
by
shayonj
6mo ago
|
0 comments
19.
▲
by
shayonj
6mo ago
That's a good shout. I will update/port it back from here - https://github.com/cloud-hypervisor/cloud-hypervisor/pull/78... , but quite fast
20.
▲
Linux Page Faults, MMAP, and userfaultfd for fast sandbox boot times
(shayon.dev)
16 points
by
shayonj
6mo ago
|
3 comments
21.
▲
Creating virtual block devices with ublk
(jpospisil.com)
2 points
by
shayonj
6mo ago
|
0 comments
22.
▲
A simple L7 proxy for vLLM that manages LoRA adapter storage via NVMes
(github.com)
2 points
by
shayonj
6mo ago
|
0 comments
23.
▲
by
shayonj
7mo ago
This is going to be an interesting space to watch I think and big part of offering sandbox as a service basically for enterprise and saas needs.
24.
▲
by
shayonj
7mo ago
Yeah, it's hard to hit the right balance with nuance around these and you're spot on. What I meant to get at was the specific difference in default modes where gVisor's systrap intercepts syscalls via seccomp traps and handle
25.
▲
by
shayonj
7mo ago
Heya! nice to see you here. In retrospect it feels like CI companies and environments are very well suited for sandboxes since a lot of the problems overlap around ephemeral workloads, running untrusted code, fast cold starts, multi-tenancy
26.
▲
by
shayonj
7mo ago
It touches in the gvisor section around the trade-off that the surface area for gvisor is smaller. There are trade offs. It’s not dishonest.
27.
▲
by
shayonj
7mo ago
That’s a good shout! I have been curious as well and did some experiments. Also left out GPU sandboxing from the post as well. Maybe will reflect in a part II post.
28.
▲
by
shayonj
7mo ago
Wasmer looks v cool. I must check it out
29.
▲
by
shayonj
7mo ago
That is a good call out and I missed to consider the options you pointed. When I am back on keyboard I will add an updated note with a link to your comment. Thank you!
30.
▲
Let's discuss sandbox isolation
(shayon.dev)
173 points
by
shayonj
7mo ago
|
71 comments
More ›