▤ Stackers Network

January 10, 2026

Stackers Network Digest — January 10, 2026

17 threads · 38 messages · openstack-discuss

The Big Picture

The first full working week of 2026 was quiet and operator-flavored, with the community easing back from the holidays — the TC explicitly skipped its January 6 meeting because members were still on PTO. The 2026.2 release-name poll closed mid-week. Beneath the surface, three forward-looking conversations got started: a renewed debate about what comes after uWSGI, a concrete proposal to add a NetApp FlexClone Nova image backend for fast NFS provisioning, and a continued operator dig into the openstacksdk memory-leak diagnosis from the prior week.

Releases & Announcements

The 2026.2 release-name poll — for names starting with "H" — ran by Allison Price and closed Thursday, January 8 at 17:00 UTC. Élod Illés posted the release countdowns for weeks R-12 and R-11 as the 2026.1 "Gazpacho" cycle progresses, and the TC confirmed there was no weekly IRC meeting on January 6, with the next one set for January 13.

Development & Technical Decisions

The openstacksdk / keystoneauth1 memory-leak conversation continues. Sam Clippinger's diagnosis from late December — that the Prometheus path inside openstacksdk.Proxy.request() records per-URL counters and histograms with no off switch, driving steady memory growth in busy services like nova-scheduler (Caracal workers reaching ~8 GB and getting OOM-killed in a 500K-VM/week build farm) — drew a follow-up thread from Mike Lowe asking about "Memory leak in nova-scheduler, possibly all services?". Operators running large fleets should track both threads; the open design question is whether keystoneauth1's Adapter/CloudRegion should honor a collect-timing-style off switch.

What replaces uWSGI? A contributor revived the long-running question of what should follow uWSGI, noting it has now been in maintenance mode for three years and that the openstack-ansible plan from the Epoxy PTG to find a replacement hasn't materialized — OSA still relies on uWSGI. The open question to the community: what should the next "standard" WSGI runtime be — gunicorn, uvicorn, granian, or something else — given operators benefit most from a single well-documented option. No decision yet, but a real strategic gap.

NetApp FlexClone for Nova ephemeral storage. Jayaanand Borra proposed a new NetAppNfs image backend in nova/virt/libvirt/imagebackend.py that calls NetApp's ONTAP REST /storage/file/clone endpoint to offload cloning to the array — aiming for Ceph-like, sub-second, metadata-only instance creation on NFS instead of the slower QCOW2-over-NFS copy path. He's soliciting design feedback on driver red flags, credential handling via oslo.privsep, and community interest. A plausible performance win for NFS-backed deployments if it gains traction.

Smaller technical threads. Sylvain followed up on the Designate "Updating a root zone NS record is not allowed" restriction blocking a multi-region PRIMARY/SECONDARY setup across three independent deployments, asking whether the ~9–10-year-old restriction could be made configurable (via policy or a flag). A Nova resize bug with swap vs. no-swap flavor combos — producing a duplicate vdb libvirt XML error — drew confirmation and a recommendation to file the bug. A Lajos Katona thread on Keystone interpreting expired LDAP passwords asked specifically what process (spec, RFE, or bug) would be needed to add the feature.

Heads Up / Action Needed

  • OSA Galera backup change. keshav bareja flagged that after upgrading openstack-ansible to Epoxy (2025.1), Galera moved to MariaDB 11.4.8 and the old mysqldump --all-databases no longer works; he's asking whether mariadb-dump --all-databases --single-transaction --quick --skip-lock-tables ... is the right replacement. Worth answering for any OSA Epoxy operator.
  • vmware_nsxv Kolla deployments are broken. Parsa Aminian reported neutron_plugin_agent: "vmware_nsxv" containers exiting with non-zero on Kolla-Ansible — a thread to watch if you run that combination.
  • Acknowledgment of a Puppet OpenStack departure. Takashi Kajinami publicly thanked Alan Bishop for his work on the puppet-cinder/glance/manila cores after Bishop's resignation announced the prior week.

Community & Events

Jay Faulkner opened a pointed critique of the new metrics dashboards as a replacement for the existing (Bitergia-style) ones. His three core use cases — per-reviewer Gerrit stats (review types, frequency, reviewer comparison), contribution data beyond the top-5 companies, and per-OpenStack-project (not just per-repo) filtering — feel poorly served by the new tool. This matters because Bitergia is being wound down and several reviewer-evaluation workflows depend on it.

Kostya Velychkovsky pointed a Cinder snapshot-lifecycle thread participant toward OpenStack Freezer as a built-in option overlapping the goals of last week's SnapSentry tool. Tacker cancelled its weekly IRC meeting again. Otherwise the list returned to its usual mix of operator how-tos — vmware_nsxv breakage on Kolla and a Ceilometer/Gnocchi question about creating metrics for list fields in event payloads — as the community geared back up for the second half of the Gazpacho cycle.