Operations on a VM now write their rows the way the admin audit does: one
row per operation, done or failed with a safe reason, with the user, IP and
host. Each row names the VM ("VM web-1: delete snapshot snap1") and carries
its UUID in the new Logs.instance_uuid column.
- audit_vm and failed_vm wrap audit() and failed() for a VM. A failure the
view turns into a message is caught outside the block, so its failed row
is written; busy claims, permission and quota refusals and invalid input
write nothing. No password, key or XML goes into a row.
- All VM page operations move from addlogmsg to audit_vm, and addlogmsg
goes. QoS changes and defining a VM from XML are logged now; a gstfsd
error reply (root password, SSH key) is a failed row, not a done one.
- Partial results are rows of their own: the VM delete, then each disk
delete; the disk detach, then the volume delete; the migration, then a
failed autostart restore, NVRAM removal or record on the target.
- Migration rows use the compute name as the host column, as every other
row does.
- VM job rows (create, clone, snapshots) and their resolutions carry the
VM's UUID; snapshot job intents now record it. The owner rows of the VM
page and the accounts pages, and the ownership system rows, carry it too.
- The Logs tab loads vm_logs/<pk>/ and shows the rows with the VM's UUID,
so a rename keeps the history, two VMs with one name are told apart and
a VM on two computes is one VM. It gains a Result column. Rows from
before this change have no UUID and stay on the Logs page only.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The machine types, volumes, templates and disk buses load when their
parent changes. A slow answer for an earlier choice could arrive last and
fill the list for the newer one, and while a list loaded it still held
the previous choice's items, which could be submitted.
Each list now empties when its load starts, and a newer load aborts the
older request, so only the answer for the current choice fills it. Chosen
volumes stay: they are found by name in any pool.
Fixes#177
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The dialog looped over every compute and skipped the unreachable ones,
so its "No compute is reachable." message showed only when there were no
computes at all. With hosts that were all offline it was an empty list.
The index view now gives the dialog the reachable computes, and the
template lists those, so the message covers that case.
Fixes#176
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
On the machine step, "Or define it from XML" lost the definition when it
was refused: a taken name re-rendered the page with an empty, collapsed
editor; malformed XML went on to libvirt, and libvirt's refusal sent the
browser back to an empty page with the reason at the top.
All three refusals now render the page with the section open, the
submitted XML in the editor (and in the field it posts, so it can be sent
again unchanged), and the reason above the editor, which takes the focus.
Fixes#174
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The create page's flavor dialog posted its forms the normal way, and the
server sent a fresh page back: the VM's name, sizes, disk choices and MAC
edits were lost, and a refused flavor left the page for a bare form.
The dialog's forms are now posted from the page. The flavor views answer
such a post with the presets and the dialog body, rendered from the same
includes as the page, and the script swaps them in: a refused flavor shows
its errors in the dialog, and the focus stays in it. A post without the
script, or one whose request fails, still goes the normal way.
Fixes#173
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The create form and the API took 0 or a negative number for vCPUs,
memory and a new disk's size: the API queued the job (202) and it
failed later in libvirt; the page did the same, its formnovalidate
skipping the browser's min. The create model's fields now refuse
values below 1, so the form shows the error beside its field and the
API answers 400 naming it.
The form also checks the MAC addresses as the API and the job do (one
per NIC at most, each a MAC address), so a bad one is an error on the
form instead of a job that fails.
Closes#175
The live suite run against v3.1.0 on two Vagrant computes (TCP and SSH,
TLS, a socket inside compute1) found two tests that no longer held, and
nothing else:
- A SATA disk added to a running VM is refused by libvirt; since
90c148d a refused form goes back to its page with the reason instead
of a 500. The test expected the 500; it now expects the redirect and
the message, and still checks that neither definition changed.
- The NVRAM copy test built its libvirtError("define failed") and
matched that text, but libvirt-python gives a new libvirtError the
message of the last error its thread recorded: in the socket run a
connection of an earlier live test was closed by then ("client socket
is closed"). The test now checks that its own error object is raised.
The create page had three tabs (flavor, custom, template), each with its
own copy of the same create form; every flavor rendered one more copy in
a dialog, and fixes had to be made in each copy. All of them posted the
same form to the same job, differing only in where the disk comes from.
The page is now one form in sections: name and size, disk, networks and
an Advanced section. The disk section offers three sources as radio
cards (a new disk, existing volumes, a copy of a template); only the
chosen source's fields are shown, and the others are disabled, so they
are neither sent nor validated. A template's copy goes into the
template's pool unless another is picked. Flavors are presets that fill
vCPU, memory and disk size and pick a new disk; they are added and
deleted in a dialog outside the form. Every source uses the
per-interface network rows, so a new disk no longer requires a MAC:
libvirt gives an interface without one its address, as it did for the
other sources. A refused form comes back with its source, volumes,
template and networks, and the focus on its first error. Without a
storage pool or a network the page shows why instead of the form, and a
host that cannot be read is a message, not a 500.
The host dialog and the machine page share the form's look: hosts are
rows with their CPU and memory use, the machine page is a card with
Cancel/Next and the XML way is a collapsible section. All three steps
show the same step indicator, done steps linking back.
The create job refuses more than one disk source (the page sends one; a
crafted request could send two, and the job would have used the first).
A field with help text is described by its help and its error.
Frontend review fixes: the presets and the disclosure buttons show a
focus ring for keyboard users; a disclosure spans its card and its
chevron turns while open; a preset says it picks a new disk; numeric
fields bring a numeric keyboard and MAC fields no autocorrection; the
machine page's Next waits for a machine type. Tom Select's control
keeps the text colour of the select it replaces: its own fixed colour
made typed text invisible on dark Bootswatch themes, on every page that
uses it.
The Logs page now records the panel's own records too: users (add,
change with the changed fields, delete, block, unblock, an admin's
password set, a user's own password change including the first-login
one, own profile edits), groups with their members and permissions,
VM owners from the accounts pages and from the VM page, OTP QR code
mails, flavors from web and API, app settings and the theme,
removed-VM records and SSH keys. No row carries a password or a key.
- A user is created in one transaction: when a VM named in
NEW_USER_DEFAULT_INSTANCES is missing, no half-made user stays.
- The OTP mail checks the email address before it creates a device;
without one it writes a failed row and makes no device.
- The VM page's owner rows read "VM web-1 owner bob: add" and
"...: remove" instead of "Add owner: bob" and "Delete owner: 17"; a
second owner refused by the single-owner setting is a failed row.
Owner rows keep the VM's host and name, so the VM's log tab shows
them.
- A setting or theme posted with its current value is no change and
writes no row; the old free-text setting rows are gone.
Host-side objects (networks, storage, interfaces, secrets, filters)
follow in the next layer.
The Logs page shows who added, changed or deleted a compute, from web
or API, and every login, failed login, lockout and logout, with the
client address and a Result column (Done, Failed, Unknown for older
rows). A failed change is recorded with its reason and the error still
reaches the user; a refused one is recorded too.
- logs.audit.write saves each row in its own savepoint and never
raises, so a log write cannot break or undo the action it records.
The audit() context manager writes one row when the change ends.
- Logs gains success and ip; user, host and instance are wider and
date is indexed. The page orders by date then id.
- A compute edit names the changed fields; an edit that changes
nothing writes no row, and an empty password field is no change.
- A compute's VMs leaving the panel get a system row each: owners
kept on the Removed VMs page, moved to another compute, restored,
left without owners, or the removed record expired.
- Job rows carry the result. A job's log entry that cannot be
computed is logged and skipped instead of rolling back the finish
(test_service.py changes this on purpose).
- Logins are recorded by signal receivers; CustomLoginView is removed
and the plain LoginView is used. A wrong OTP code is a failed login,
counted by django-axes toward the lockout.
Panel records and host-side objects follow in the next layers.
In the VM lists, Force Off and Power Cycle ran on a single click,
though both cut the VM off as a power failure does. Each button now
asks first, naming the VM and saying that what the guest has not
written to disk is lost.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
No page or script called it, and it answered with the disk device
types for both videos and buses. The create view keeps reading the
domain capabilities through wvmConnect.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Three selects of the VM create page had no accessible name. TomSelect
points the label of a select it enhances at its own input, so the
network select lost its label and the image select never had one; the
template select sits beside the template pool's label with none of its
own. Each now has an aria-label (Images, Networks, Template), which
TomSelect also copies to its input for the image select.
A refused custom create form came back with its image pool unselected
and its disk and network lists empty: every disk's device and bus and
every MAC had to be chosen again. Adding a disk or a network also reset
the rows already there: each addition rebuilt all rows from an HTML
string, so the earlier disks fell back to the default device and bus
and every NIC got a new random MAC, a typed one included.
The view sends the submitted pool, disks (volume, device, bus) and
networks (name, MAC) back with json_script, built field by field from
the POST (submitted_custom_lists). The rows are cloned from two
<template> elements and filled with textContent and listeners; one
function, sync_custom_lists, derives the images, networks and mac
fields and the device<i>/bus<i> names from the rows in order. Adding or
removing a row leaves the others as they are, and a random MAC that
arrives after one was typed is dropped. On load the page selects the
pool, waits for its volumes, rebuilds each disk with its device, waits
for that device's bus list and selects its bus, rebuilds the networks
with their MACs, and then focuses the first error. A submitted value the
loaded lists no longer have stays selected. The bus and volume loaders
return their requests so that the page can wait on them.
When the server refused a VM create submission (a field error, a busy
resource, a reused submission key, an existing VM name, a missing MAC
for a new disk), the page came back on its first tab with the flavor
dialog closed and every entered value gone: the hand-written fields
showed fixed defaults, and the errors sat in a block at the top of the
page, outside #messages and tied to no field.
The view builds one NewVMForm per create form on the page (custom,
template, one per flavor; create_forms), each with its own defaults; a
hidden form_id, which is also the prefix of the form's field ids, names
the submitted one, and it comes back bound. Every field renders from its
own form and is marked by two small template tags (form_fields:
error_attrs, field_error) when refused; a summary at the top of the
form lists all of its errors, also those of fields it keeps hidden. The
submitted tab or flavor dialog, marked data-form-id, reopens with any
"Advanced" section that holds an error; a reopened dialog shows the
messages too; the template form gets its pool, template and storage
back. base.html's error focus is a function that walks the places an
error can be, and a page can call it once it is set up, as the create
page does. form_id changes nothing in how a POST is processed; one of
an unknown form (a flavor deleted since) has its errors in the
messages. The default video now applies to every create form.
The custom form's disks and networks are not restored yet.
The custom create form shows a MAC per network (random, or typed in),
but had no mac field: the script wrote the joined list into every
name="mac" input of the page, the other six-plus forms' own MAC fields,
so a flavor form then sent one MAC per custom network and was refused,
and the custom VM got other MACs from libvirt. The custom form gets its
own hidden mac field and the script writes only there. The field took
17 characters, one MAC, so the API also refused a VM with two NICs and
their MACs; it takes 255. The serializer's test sent no submission_key,
so its check for more MACs than networks never ran; it does now, next
to the case of two networks with two MACs.
Changing a disk's device type in the custom create form asked for its
buses at /archs/undefined/machines/undefined/...: the script read the
architecture and machine from a select[name=arch] and #machine-control
the page does not have. The request failed (500) and the bus list kept
the disk's buses, so a CD-ROM could be put on virtio. The script takes
both from the page's own URL.
The grouped list opens one tbody per compute, so with no compute the
sortable table had none, and sortable.js threw on page load ("Cannot read
properties of undefined (reading 'rows')"). An empty tbody is rendered
in that case.
The Vagrant computes serve a 1 GiB LUN on 127.0.0.1 (targetcli). A live
test, run with TEST_LIBVIRT_ISCSI_TARGET, starts a transient iSCSI pool
on it and attaches the LUN from the VM page to a stopped and a running
VM: both definitions get a raw block disk on its /dev/disk/by-path link.
libvirt's logical pool cannot change a volume's capacity, so every
resize of such a disk ended in a 500. Disks are resized one at a time;
VIR_ERR_NO_SUPPORT becomes a message naming the disk, and the other
disks still grow.
A live test creates an LVM pool from the storage page on a spare disk
(TEST_LIBVIRT_LVM_DEVICE; the Vagrant computes get /dev/vdb), attaches
new and existing volumes to stopped and running VMs as block disks, and
checks the refused resize in every state.
Settings > Console gets a Console type field. Serial removes the VM's
graphics device (adding a pty console when there is none); VNC adds the
graphics device new VMs get, on localhost, and then applies the form's
VNC fields, so an administrator can open it to all interfaces for a TCP
compute. Serial is offered only while SERIAL_CONSOLE_ENABLED is on.
The xterm console never connected: socketiod expected a different token
format, its worker had no compute credentials, one global console was
shared by every client and its output broadcast to all of them, and the
Docker image did not run socketiod at all.
- console/tokens.py: one authorize() for noVNC and socketiod; the signed
token is the only credential, checked at connect time.
- socketiod: one pty worker per client (consolecallback with the
instance id), one session per VM, a session cap, input and resize
checked; the parent never opens libvirt. The engine.io same-origin
check is on.
- consolecallback: opens the console with the compute's credentials and
exits with a message when it cannot.
- Docker runs socketiod as a runit service; nginx proxies /socket.io/
with Host and the scheme for the origin check. Supervisor and systemd
start it as well; WEBVIRTCLOUD_SERIAL_CONSOLE=1 enables it.
- Tests: token and session unit tests, socketiod as a process, and a
live test that reads SeaBIOS from a wvc-test VM on the TCP and SSH
hosts.
The move copies the VM's NVRAM file to the destination with libvirt
streams (unless both hosts share the NVRAM directory) and removes the
copy if the move fails; the source's file goes with the existing
cleanup. The undefine keeps the file, as libvirt refuses to undefine a
VM with NVRAM otherwise. A VM with TPM state is still refused.
libvirt's backup copies the disks of a running or paused VM from one
instant into new volumes while it keeps running; the clone is defined
shut off as before. The VM is held for the copy, the copy can be
cancelled, and recovery reads the job's records and libvirt.
A disk-only snapshot was taken without any space check, though its
overlays grow as the guest writes and a later delete merges them down;
a full pool pauses the guest and can leave the snapshot undeletable.
Only a memory snapshot was checked, for the memory alone.
Before a snapshot, each storage pool with a disk of the VM must have
free a share of those disks' virtual size (10%, at least 1 GiB), and
the pool of the memory file the VM's memory on top. Both values are on
the Settings page (Snapshot Free Percent, Snapshot Free Min GiB; 0
turns a rule off, an invalid value keeps the default). The pools are
refreshed first, and a failed refresh or an unreadable volume refuses
the snapshot instead of skipping the check; a disk in no pool is not
checked. Nothing is reserved: later writes can still fill a pool.
A disk-only snapshot of a running VM is crash-consistent: data the
guest has not written yet is missing from it. With a connected QEMU
guest agent, the snapshot form now offers to freeze the guest's file
systems while the disks are snapshotted (libvirt's QUIESCE flag); the
box is checked by default and can be cleared. It is offered only for a
running VM whose agent is connected, and not together with memory:
libvirt accepts quiesce only for disk-only snapshots, and a memory
snapshot keeps the whole guest state anyway. A request that does not
fit is refused before libvirt is called.
Reverting a running or paused VM to a disk-only external snapshot
powers it off: libvirt starts it again from the snapshot's disks, as
after a power cut. The confirmation now says so for such a snapshot; a
snapshot with memory resumes the VM as it was saved.
libvirt deletes any volume it is asked to. The storage page and the API
deleted a VM's disk, the base image of another volume or the file a
snapshot was taken on without a check, breaking the VM or its
snapshots.
wvmStorage.del_volume now refuses a volume a domain on the compute uses
as a disk or a backing file of one, a file of a domain's snapshots, or
the backing file of another volume. The check is the one destroy
already used, moved to wvmConnect and extended to snapshot files, so
destroy also keeps the files of another VM's snapshots. The storage
page shows the refusal; the API answers 400 with it, as for other
refusals. Deleting a disk from a VM refuses after the detach when the
VM's own snapshots need the volume, and says the volume was kept.
Network disks count by their source name, pool volume disks by their
volume's path. A volume whose metadata cannot be read stops the check
instead of counting as unused; only one removed meanwhile is skipped.
The check and the delete hold the compute lock, so no VM operation
attaches the volume in between; a busy compute is a message on the page
and 409 in the API.
Deleting a VM with its data deleted only the images its disks pointed
at. With external snapshots that is the newest overlay: the image the
snapshots were taken on, the older overlays and the memory files stayed
on disk, with nothing referring to them once the VM and its snapshot
metadata were gone.
The files the VM's snapshots were taken on, down each disk's backing
chain, the overlays of snapshots kept off the chain after a revert, and
the snapshots' memory files are now the VM's own too. A template the
VM grew from ends the chain and stays, as does an overlay of a disk
detached since. The backing references of the VM's own volumes no
longer count as another domain's use, and a file another VM uses keeps
the files below it.
A backing file that cannot be read (other than one in no pool) stops the
split before anything is deleted, instead of ending the chain there.
The snapshot tab listed the VM's snapshots three times per page: once
per kind and once more for all of them. The list of all is now made
from the two kinds, so the combined flag is no longer used anywhere.
The revert confirm named only the disks the persistent definition
would lose. A disk hot-plugged into the running VM alone is detached
too, so the live definition is compared with what ran when the
snapshot was taken.
A storage pool refresh error after libvirt had already made, reverted
or deleted the snapshot was shown as if the operation had failed. It
is now logged, and the left-behind check is skipped then, since a stale
pool still lists the removed files. A volume removed while that check
reads it is skipped too.
An error reading the block jobs no longer breaks the instance page;
create, revert and delete still query the jobs again and refuse.
When libvirt cannot remove a snapshot file it merged or discarded, it
only logs a warning and the delete or revert still succeeds. On NFS it
removes them as the qemu user, so an image directory the qemu user may
not write to left every merged overlay on disk without a word.
Revert and delete now compare the files in use before and after and
show a warning naming the ones still on disk. Only a snapshot's own
files (overlays, memory) and overlays on a snapshot point count, and a
disk the revert detached is never named, so a kept disk is not offered
for removal.
Create, revert and delete are one libvirt call each. The pflash-to-rom
loader rewrite, the hand-written block commit and the s1./s2. names are
gone, so UEFI VMs can take external snapshots, a revert keeps disks added
later, and deleting an older snapshot keeps the newer one.
Create refuses what libvirt could not revert or delete later: internal
snapshots on the VM, a saved state, a running snapshot job, disks that
are not files, CD-ROM media, a branch after a revert, and memory without
a running VM or room in the first disk's pool. A disk-only revert keeps
the VM running, paused or shut off; one with memory resumes it as saved.
A revert in the same second as the previous one, which libvirt fails
after stopping the VM, is retried once.
A snapshot the old code took of a UEFI VM saved the loader as rom; its
revert is refused (deleting it still works).
The snapshots tab lists internal and external snapshots separately, can
save memory too, shows running snapshot jobs, and the revert confirmation
names the disks it detaches and that NVRAM is not restored.
131 test modules sat next to the code they test; instances alone had
51 of them beside views.py and utils.py. They move into a
tests/ package per app, and the JavaScript checks into
webvirtcloud/tests/js. Each non-empty tests.py becomes tests/test_views.py;
the empty generated stubs are removed.
Imports of the app's own modules become absolute, paths built from
__file__ go one level up, the live helper is instances.tests.livetest,
and coverage omits */tests/*. The suite discovers the same 960 tests as
before.
