# mvirt 0.24.11

migrations: no
wire: unchanged
rev: 41f8d8d0684ab927ef9f1a72cf2a06852a0ecf9c

mvirt 0.24.11

A patch release: the last step that held a new leader's first slice push.

What changed

- #341: restore-point scheduling and retention (ADR-0058 D4) run as their own
  leader-scoped task every 30 s instead of as a converger ladder step. The
  ladder timing shipped in 0.24.10 named it: its retention deletes expired
  restore points from the repository one object at a time, and it held a
  fresh leader's first push for 66 s (7 min on the previous roll, with the
  object gateway restarting). Together with 0.24.9's changes a leadership
  change should now hand every node its slice within seconds.

Upgrade: ordinary host-by-host roll (release.md §C1 + §C3). No store migration.


## Upgrading

Point your flake's `mvirt` input at this release and redeploy the control-plane
host first, then each node in turn:

```
nix run .#colmena -- apply boot --reboot --on <control-plane host>
nix run .#colmena -- apply boot --reboot --on <node>   # one at a time
```

This release adds no migrations, so the control-plane host rolls back the same way
a node does: boot the previous NixOS generation.