Repository navigation

Issues

New issue

Dependency Dashboard
#4239 · oxide-renovate opened on Oct 10, 2023
Tracking Issue for "zones can run out of resources"
#7468 · smklein opened on Feb 3, 2025
4

some of reconfigurator-cli's sled state management is error-prone and not faithful to real systems
oxidecomputer/omicron#11592 · opened on Oct 23, 2026
1
Racks need to be (soft) deletable
oxidecomputer/omicron#11583 · opened on Oct 23, 2026
want panics in integration_tests::unauthorized::test_unauthorized to include the endpoint and request
oxidecomputer/omicron#11582 · opened on Oct 23, 2026
Prevent Sled-Agent upsert until rack is adopted by operator
oxidecomputer/omicron#11580 · opened on Oct 23, 2026
Add a crdb partial unique index for primary NICs
oxidecomputer/omicron#11568 · opened on Oct 22, 2026
Instance state change gets 404 during start saga; Nexus marks VMM as Failed but propolis keeps running.
Bug
oxidecomputer/omicron#11560 · opened on Oct 22, 2026
5
Multirack Join Completion response should return the bootstore generation number
oxidecomputer/omicron#11549 · opened on Oct 21, 2026
1
Guests running fstrim on a local disk may trigger extreme I/O latency for that drive
oxidecomputer/omicron#11543 · opened on Oct 21, 2026
3
Mechanism to stop new IP allocations in an IP pool
Enhancement
oxidecomputer/omicron#11542 · opened on Oct 21, 2026
3
Plumb "size limit" through support bundle API
oxidecomputer/omicron#11541 · opened on Oct 20, 2026
If we hit a support-bundle size limit, we should be able to identify that in the API
oxidecomputer/omicron#11540 · opened on Oct 20, 2026
Add SP selector for support bundles
oxidecomputer/omicron#11531 · opened on Oct 20, 2026
fm: want fm_analysis input loading to be in parallel
oxidecomputer/omicron#11523 · opened on Oct 17, 2026
2
Give nexus ability to complete the multirack join process
oxidecomputer/omicron#11521 · opened on Oct 16, 2026
Sled-agent API for exposing found DDM peer sled-agents/racks
oxidecomputer/omicron#11513 · opened on Oct 16, 2026
tofino-monitor shutdown of switch zone can race with sled-agent booting it
oxidecomputer/omicron#11512 · opened on Oct 16, 2026
12
/system/hardware APIs that include locations from inventory data should probably say when the inventory was collected somehow
Enhancement
oxidecomputer/omicron#11509 · opened on Oct 15, 2026
Ensure that the rack network config for a newly joined rack is on the same /48.
Enhancement
oxidecomputer/omicron#11508 · opened on Oct 15, 2026
Bootstore configuration needs to take racks into account.
Bug
oxidecomputer/omicron#11507 · opened on Oct 15, 2026
RSS -> Nexus handoff fails if a scrimlet doesn't realize it's a scrimlet early enough
oxidecomputer/omicron#11506 · opened on Oct 15, 2026
Use an enum to mark that a VMM should be stopped
oxidecomputer/omicron#11490 · opened on Oct 14, 2026
4
Multirack Join Completion response is a little strange on cold boot
Enhancement
oxidecomputer/omicron#11477 · opened on Oct 12, 2026
Modular Router Configuration (RFD 662) - Tracking Issue
Enhancement
oxidecomputer/omicron#11457 · opened on Oct 8, 2026 · 24
PortTicket::release() appears to leak memory
oxidecomputer/omicron#11450 · opened on Oct 7, 2026
could we raise ereports when oxide processes call fork (etc) instead of posix_spawn?
oxidecomputer/omicron#11445 · opened on Oct 7, 2026
3

Navigation Menu

Repository navigation

Issues

Issues

some of reconfigurator-cli's sled state management is error-prone and not faithful to real systems

Racks need to be (soft) deletable

want panics in `integration_tests::unauthorized::test_unauthorized` to include the endpoint and request

Prevent Sled-Agent upsert until rack is adopted by operator

Add a crdb partial unique index for primary NICs

Instance state change gets 404 during start saga; Nexus marks VMM as Failed but propolis keeps running.

Multirack Join Completion response should return the bootstore generation number

Guests running `fstrim` on a local disk may trigger extreme I/O latency for that drive

Mechanism to stop new IP allocations in an IP pool

Plumb "size limit" through support bundle API

If we hit a support-bundle size limit, we should be able to identify that in the API

Add SP selector for support bundles

fm: want fm_analysis input loading to be in parallel

Give nexus ability to complete the multirack join process

Sled-agent API for exposing found DDM peer sled-agents/racks

tofino-monitor shutdown of switch zone can race with sled-agent booting it

`/system/hardware` APIs that include locations from inventory data should probably say when the inventory was collected somehow

Ensure that the rack network config for a newly joined rack is on the same /48.

Bootstore configuration needs to take racks into account.

RSS -> Nexus handoff fails if a scrimlet doesn't realize it's a scrimlet early enough

Use an enum to mark that a VMM should be stopped

Multirack Join Completion response is a little strange on cold boot

Modular Router Configuration (RFD 662) - Tracking Issue

`PortTicket::release()` appears to leak memory

could we raise ereports when oxide processes call `fork` (etc) instead of `posix_spawn`?