jacobwg

25 ms·

1.
Computers Compared(fly.io)
2 points by jacobwg 28d ago|1 comments
2.
by jacobwg 1mo ago
It's to prevent theft by the delivery person or by the wrong recipient
3.
by jacobwg 1mo ago
We've been working on this at Depot[0]. We built the internals of Depot CI as a general-purpose workflow engine that understands GitHub Actions as an input (with plans to extend) - we've been working to make Actions runners more
4.
Amazon S3 Annotations(aws.amazon.com)
5 points by jacobwg 3mo ago|0 comments
5.
by jacobwg 6mo ago
Hey, I'm one of Depot's founders - we built our own custom parser for the YAML syntax that translates to Depot CI's orchestration APIs. This API will actually be a public API shortly, which will unlock some interesting use-ca
6.
Depot (YC W23) Is Hiring an Enterprise Support Engineer (Remote/US)(ycombinator.com)
1 points by jacobwg 9mo ago
7.
Depot (YC W23) Is Hiring a Staff Infrastructure Engineer(ycombinator.com)
1 points by jacobwg 10mo ago
8.
by jacobwg 11mo ago
You all should add EC2 - extra bonus if you have some way of tracking performance in addition to errors (right now we're seeing EC2 instances in us-east-1c not transition out of Pending status).
9.
by jacobwg 11mo ago
We've been observing EC2 instances launched in us-east-1c (use1-az2) remain in Pending status for a very long time / indefinitely, starting at around 16:00 UTC.
10.
Depot (YC W23) Is Hiring a Principal Design Engineer (Remote US/EU)(ycombinator.com)
1 points by jacobwg 1y ago
11.
Discovering useful third-party GitHub Actions(depot.dev)
4 points by jacobwg 1y ago|0 comments
12.
Depot (YC W23) Is Hiring a Community and Events Manager (Remote)(ycombinator.com)
1 points by jacobwg 1y ago
13.
by jacobwg 1y ago
Since we launched this last year, GitHub released a v2 of their internal cache API [0], based on Twirp [1] of all things, so we adapted to that. Interestingly that Twirp service also receives Actions artifacts, though we have not intercepte
14.
Depot (YC W23) Is Hiring a Technical Content Writer (Remote)(ycombinator.com)
1 points by jacobwg 1y ago
15.
Flox Build and Publish(flox.dev)
2 points by jacobwg 1y ago|0 comments
16.
Depot (YC W23) is hiring an enterprise support engineer (UK/EU)(ycombinator.com)
1 points by jacobwg 1y ago
17.
by jacobwg 1y ago
Depot (W23) | Enterprise Support Engineer | Remote (UK, Europe) | Full-time | €100K - €160K | https://depot.dev Depot is the fastest place to build software. We accelerate builds for customers using GitHub Actions, Docker, Bazel
18.
by jacobwg 1y ago
Not quite for every container, but we operate a multi-tenant remote build execution service (container builds, GitHub Actions jobs, etc) so we launch a lot of ephemeral VMs in response to customer build requests. We use separate EC2 instanc
19.
by jacobwg 1y ago
We also do GitHub Actions runners as a service, so a very high volume of differently-sized ephemeral VMs. We’ve experimented with .metal hosts, however they represent a bin-packing optimization problem, in that you will always be running so
20.
by jacobwg 1y ago
Yeah we do some similar tricks with our registry[0]: pushes and pulls from inside AWS are served directly from AWS for maximum performance and no data transfer cost. Then when the client is outside AWS, we redirect all that to Tigris[1], al
21.
by jacobwg 1y ago
No worries, entirely valid question. There may be ways to tune page cache to be more like this, but my mental model for what we've done is effectively make reads and writes transparently redirect to the equivalent of a tmpfs, up to a c
22.
by jacobwg 1y ago
Exactly, a ramdisk-backed writeback cache for the root volume for Linux. For macOS we wrote a custom nbd filter to achieve the same thing.
23.
by jacobwg 1y ago
Today we (Depot) are not, though some of our customers configure this. For the moment at least, the ephemeral public IP architecture makes it generally unnecessary from a rate-limit perspective. From a performance / efficiency perspect
24.
by jacobwg 1y ago
The block level has two advantages: (1) you can accelerate access to everything on the whole disk (like even OS packages) and (2) everything appears as one device to the OS, meaning that build tools that want to do things like hardlink file
25.
by jacobwg 1y ago
A list of fun things we've done for CI runners to improve CI: - Configured a block-level in-memory disk accelerator / cache (fs operations at the speed of RAM!) - Benchmarked EC2 instance types (m7a is the best x86 today, m8g is t
26.
by jacobwg 1y ago
We're having some success with doing this at the block level (e.g. in-memory writeback cache).
27.
by jacobwg 1y ago
I'd love to experiment with that and/or flags like `noatime`, especially when CI nodes are single-use and ephemeral.
28.
Disk I/O bottlenecks in GitHub Actions(depot.dev)
106 points by jacobwg 1y ago|72 comments
29.
A Shell for the Container Age: Introducing Dagger Shell(dagger.io)
4 points by jacobwg 1y ago|0 comments
30.
Depot (YC W23) is hiring a founding developer marketer (EU/US remote)(ycombinator.com)
1 points by jacobwg 2y ago

More ›