/api/storage now runs a single `zpool status -j --json-int -P` (OpenZFS
2.3+) for all pools and returns, per pool:
- top-level vdevs with type, raidz/draid parity, class (data, special,
dedup, log, cache, spare), state and read/write/checksum errors
- member devices with path, resolved base disk, state, error counters,
slow I/Os, and the spare-N/replacing-N they sit under while swapping
- scrub/resilver status (progress only while running), the pool's
status/action text and data-error count
- fragmentation, capacity and dedup ratio from `zpool list`
This replaces the per-pool `zpool status` text loop whose result
(zfs_disks) was never used. Member paths (by-id, by-partlabel, wwn) are
resolved to base disks via sysfs rather than an lsblk call per device.
On ZFS without `-j`, pools report details_available: false and the UI
falls back to the existing basic card.
The ZFS pool card shows a layout summary (e.g. "RAIDZ2 | 5 disks |
Scrub finished ...") and opens a details dialog with the vdev tree,
per-device error counts, scrub/resilver progress, and links from each
member disk to its disk details dialog.
Tested against `zpool status -j` output from PVE 9 / OpenZFS 2.4.4:
mirror, raidz2, striped mirrors, single-disk pool, and a file-backed
degraded raidz1 with log, cache and spare devices.
Tests cover the `zpool status -j` -> Storage view shape against real
output from PVE 9 / OpenZFS 2.4.4 hosts (serials and WWNs replaced):
mirror, raidz2, single-disk pool, slow I/Os, and a degraded raidz1 with
log, cache and spare devices; plus a synthetic resilver with a spare
swapping in. Member path -> base disk resolution is tested against a
stubbed sysfs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Each disk in /api/hardware now carries a controller_slot, resolved from
the disk's sysfs device path: the deepest PCI address in that path is the
controller (bridges sit above it), normalised to lspci's slot format so it
matches pci_devices[].slot directly. NVMe falls back to the controller's
own sysfs link when the namespace resolves to a subsystem with no PCI
parent; zvols and other virtual devices get None.
Hardware page:
- The disk dialog shows its storage controller (click through to the
controller), and the PCI device dialog lists the disks attached to it.
- Storage controllers in the PCI Devices list get a badge with the
number of attached disks.
Storage view:
- /api/storage returns controller_slot on each physical disk plus a
controllers map (slot -> class, vendor, device, subsystem name,
driver), built from the cached `lspci -vmm` output so the Storage view
doesn't need the heavier /api/hardware enumeration.
- "Physical Disks & SMART Status" can group disks under their storage
controller, with a header showing the controller name, slot, driver
and disk count. Groups follow PCI slot order; disks with no resolvable
controller are listed last. USB disks keep their own section.
- The disk details dialog shows the storage controller on the Overview
tab.
- Grouping is opt-in via a new Storage Display card in Settings, stored
per browser (localStorage) and synced across open views, following the
existing nav-tab-order preference pattern. Off by default.
Also:
- Treat "IDE interface" as a storage controller, so SATA controllers left
in BIOS IDE mode (e.g. Intel 7 Series, ata_piix) appear in pci_devices
and disks behind them can be mapped.
- lspci prints "Device xxxx" when it has no subsystem name; that
placeholder is ignored so the chip name is shown instead.
- Add the missing sdevice field to the PCIDevice type.
Tests cover the sysfs path -> controller slot rule against real paths
from two PVE 9 hosts (SAS HBA, AHCI SATA, SATA in BIOS IDE mode, USB
behind a hub, NVMe behind chipset bridges), the non-zero PCI domain case
(Intel VMD), the NVMe multipath fallback, and get_storage_controllers'
lspci parsing including the "Device xxxx" placeholder and driver lookup.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Draw ZFS ARC only from recorded history, leaving it out on PBS
- Read the maintenance jobs of a local PBS from its API, with their last state
- Show the OK state of PBS jobs in green whatever its case
- Compress rotated logs at once in the Log2RAM and logrotate profiles
- Offer the Log2RAM 1.8 and logrotate 1.2 update to existing installations
- Open the post-install menu from the PBS menu
- List only the customizable options that apply to PBS
- Offer the PBS no-subscription repository through the native repository API
- Upgrade PBS with the package list shown before applying it
- Remove and restore the subscription notice on the PBS web interface
- Install Figurine on PBS without the Proxmox VE aliases
- Offer the Helper-Scripts PBS post-install and microcode scripts
- Tune firewall bridges after the systemd udev rules in the automated post-install
- Add PBS repository policy tests and syntax checks
- Document the PBS differences on the post-install page
- Use backups in the Spanish PBS update message
- On a PBS host the Monitor no longer starts the node metrics and VM prewarmers, the LXC update scheduler, the managed installs detector or the app update emitter, which read guests and PVE metrics that do not exist there
- pmx_require_pve refuses only a host known to be PBS, so an unreadable package list leaves the PVE scripts working as before
- The PBS menu numbers its options, shows the tables proxmox-backup-manager prints instead of raw JSON, and says Goodbye on exit
- The PBS update asks to press Enter before returning
- Spanish: Datastores de PBS
- An explicit --devices list makes LVM open every device in it, while a plain scan opens only known volumes; the list now names only LVM physical volumes, read from udev by lsblk, so no other disk is touched
- RAID, multipath and encrypted volumes follow the disks they are built on, instead of turning the whole check off
- After each pvs, lvs or vgs run, the disks it read take their counters as the new baseline, so an unused disk can become idle
- Each host backup carries the node it belongs to, read from its source host or its name (`hostcfg-<node>`, a job of this node or its PBS backup ID); both backup lists also return the node name.
- The backup list offers This node / All nodes when some backup belongs to another node, with This node as the default, a count that follows the choice and a note with the backups not shown.
- Backups whose node cannot be read from their name are listed under All nodes.
- A local backup made on this node by a job that no longer exists is marked as Former job, read from its metadata file; backups without it, manual ones and those of other nodes are not marked.
- New texts in English and Spanish, modelled in the translation steady-state test.
- Tests for the node and the former job read from each backup, and for the cluster node membership notifications.
- An observation is never left out for its age and cannot be dismissed:
the rule that hid it after 30 days without a repetition, and the
clean-up that marked the ones of a missing device, are removed.
- The ones earlier versions left out are shown again; an error that came
back while hidden appears with its first date and its full count.
- The audit attributes recorded events to a disk by its serial number.
- The observations of a disk belong to its serial number, not to its
connector: a replaced disk starts clean, and the badge of the list
counts what the detail of the disk lists.
- An OCI container is recognised while it is still being installed, so
an image that carries Docker inside is not offered Docker.
- Links use the IPv4 address of a container and write IPv6 in brackets;
an answer without an address of its own is read again.
- The Monitor restores, when it starts, the DHCP clients of containers
that were started from it.
MacRimi
MacRimiandGitHub
DrOwlandClaude Opus 5.5
VAIO73
Jonatan Castro
Gabriel Gonzalez Pineda