Repository Issue Activity (beta)

#4987 envelope - v2 - followup: Replace OutboxMessage with Command

#5066 Add ingress ingestion API

#5334 Allow gs:// (native GCS backend) for object-store metadata

#4187 Ingress should overwite `x-restate-*` headers if present

#2972 Restatectl's warning about an unprovisioned cluster can be wrong

#5058 Unsampled trace context is not propagated to handlers

#5307 cluster feature to enable controlled sharding for workflows keys

#5097 Emit spans with the proper span kind

#5321 Graceful (drain) pause for invocations on VQueues

#5304 rocksdb-total-memory-size default (2GiB) guarantees OOM in small-footprint containers, despite the startup warning saying so

#5303 node-name defaults to container hostname, breaking restart against existing data dir in containerized single-node deployments

#5305 default-num-partitions=24 is oversharded for single-node deployments, with a measurable throughput cost

#5267 Decide default vqueue migration strategy

#5266 Decide on whether to enable scoped VOs

#5313 Ingress drops traceparent with flags `03` (W3C Level 2 random flag) and starts a new trace

#5306 Avoid duplicating eager state values in the invocation journal

#4862 schema: remove metadata v2 migration

#2786 Restatectl status --extra fails with error messages when some nodes are down

#5156 Failure detector can mark healthy peers dead after a local scheduling gap

#5282 When bilrost encoding is enabled, the restatectl tool fail to decode the schema metadata

#5265 Figure out how to re-enable orphaned jc entry cleanup asynchronously

#5214 `SELECT min(x), max(x)` over an introspection table returns a wrong `min` (silent, non-deterministic)

#5270 Partition column families are never dropped when a partition is placed away from a node

#5269 Make sure kafka-ingress drains correctly when the subscription is removed

#4548 Support deployment/service-level concurrency limits supplied at registration time