Compare commits

..

3 commits

Author SHA1 Message Date
0411c20b94 feat(data-plane): add routed SPSC stream transport
Implement namespace-owned stream rendezvous with incarnation isolation, cancellation, recovery, bounded local byte rings, and host/child actor protocols.

Add a replaceable stream transport boundary plus an Iroh ALPN adapter with framed records, backpressure, reconnect handling, and terminal propagation.

Wire stream sources and sinks through Myelin job deployment and Python bindings, with guarantee coverage across blob, namespace, job, byte-ring, and transport paths.

Sanitize nested Cargo build context in xtask so lint, Nextest, doctest, and maturin invocations retain stable fingerprints.
2026-08-23 15:51:24 +04:00
df72b4ff60 chore(test): enforce complete local test barrier
Deny workspace warnings and lint suppressions, consolidate Rust and Python coverage under cargo xtask test with 60-second per-test limits, and remove stale flaky, stateful, Docker, and orphaned test artifacts.
2026-08-23 15:51:24 +04:00
4854d85906 fix(dashboard): improve control and hardware views
Refine fleet and provisioning interactions, expand control-plane projections, and add CPU, memory, pressure, and storage telemetry for the local dashboard surfaces.
2026-08-23 15:51:24 +04:00

View file

@ -1,12 +0,0 @@
- Add connection map?
- Noisy telemetrics
- terminate cluster button
- popup should be ui-flavored not browser popup
- hardware stats seem broken
- button for selecting offers needs to be higher, so I dont have to scroll down. Maybe next to 'OFFERS' text.
- need some more information indicators for the remote nodes. Maybe stdout should be displayed on the Fleet focus view, so we can see activity?
- Actors and their stats are incredibly noisy, not sure its an easy fix. IFF the fix is obvious and easy, we need a way to fix the stats displayed for: mailbox (clears so fast its always 0), throughput (always too high-looking, ~4k msg/s but its processed maybe 2 or three) When focused, the throughput graph looks fine though.
- For telemetry specifically, a lot of this should be done by a task not an actor, actors are control flow, tasks are for continuous state-independant work (like already-negotiated streams using quic, and hardware stat samples on a fixed timer).
- Potential, future, instead of only 'Kill' we should have other buttons that match typical cloud and particularly vastai options?
- Actors on the Fleet single node focus panel need to not jump around like they do, it makes clicking them too difficult.
- We should add arrows, copy the vastai provisioning arrows, that allow you to sort the various actor bits quickly