This bug was fixed in the package neutron - 2:28.0.0-0ubuntu2
---------------
neutron (2:28.0.0-0ubuntu2) stonking; urgency=medium
* d/p/lp2150285-keep-mod-wsgi-operational.patch: Keep Apache mod_wsgi
operational for one more cycle (LP: #2150285). Upstream Neutron supports
only uWSGI for the API; under mod_wsgi the worker ID and shared start
time exposed by the uwsgi module are unavailable, so default network
segment ranges were never seeded and the OVN hash-ring workers raced
into deleting each other's freshly registered rows. When the uwsgi
module is unavailable but mod_wsgi is importable, the shim now:
- elects a stable 1-indexed worker ID per daemon-mode process via flock
files under /var/lib/neutron/lock;
- derives a host-wide start time keyed on the parent (daemon manager)
PID, recorded in a sentinel file under the same directory, so every
worker of one generation agrees on it and it advances on each restart
(it does not read the parent's /proc entry, which a non-root worker
cannot do under a confined Apache).
Adds unit tests for the shim.
* d/tests/neutron-api: Extend the autopkgtest to verify, under mod_wsgi,
that a default network segment range is seeded and that the OVN
hash-ring workers share a single start time which advances across an
apache2 restart; capture the unit journal and neutron logs if a daemon
fails to start.
-- Myles Penner <[email protected]> Wed, 10 Jun 2026 09:00:00
-0700
** Changed in: neutron (Ubuntu Stonking)
Status: Fix Committed => Fix Released
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2150285
Title:
Neutron 2026.1 with mod_wsgi packaging skips ML2 segment range
initialization, causing tenant network creation failures
To manage notifications about this bug go to:
https://bugs.launchpad.net/cloud-archive/gazpacho/+bug/2150285/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
