api2/nodes: add termproxy and vncwebsocket api calls
Even though it has nothing to do with vnc, we keep the name of the api
call for compatibility with our xtermjs client.
termproxy:
verifies that the user is allowed to open a console and starts
termproxy with the correct parameters
starts a TcpListener on "localhost:0" so that the kernel decides the
port (instead of trying to rerserving like in pve). Then it
leaves the fd open for termproxy and gives the number as port
and tells it via '--port-as-fd' that it should interpret this
as an open fd
the vncwebsocket api call checks the 'vncticket' (name for compatibility)
and connects the remote side (after an Upgrade) with a local TcpStream
connecting to the port given via WebSocket from the proxmox crate
to make sure that only the client can connect that called termproxy and
no one can connect to an arbitrary port on the host we have to include
the port in the ticket data
docs: epilog: point "Proxmox Backup" hyperlink to pbs wiki
This changes the "Proxmox Backup" hyperlink, which is referred to throughout the
Proxmox Backup Server documentation. Following this patch, it now points to the
pbs wiki page, rather than the unpublished product page.
*Note: This change is only a temporary measure, while the product page
(https://www.proxmox.com/proxmox-backup) is in development.
Stefan Reiter [Tue, 21 Jul 2020 11:41:07 +0000 (13:41 +0200)]
add .../apt/update API call
Depends on patched apt-pkg-native-rs. Changelog-URL detection is
inspired by PVE perl code for now, though marked with fixme to use 'apt
changelog' later on, if/when our repos have APT-compatible changelogs
set up.
list_installed_apt_packages iterates all packages and creates an
APTUpdateInfo with detailed information for every package matched by the
given filter Fn.
Sadly, libapt-pkg has some questionable design choices regarding their
use of 'iterators', which means quite a bit of nesting...
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Thomas Lamprecht [Thu, 23 Jul 2020 07:45:48 +0000 (09:45 +0200)]
client: rework logging upload size, bandwidth, ... info
Track reused size and chunk counts.
Log reused size and use pretty print for all sizes and bandwidth
metrics.
Calculate speed over the actually uploaded size, as else it can be
skewed really bad (showing like terabytes per second)
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Stefan Reiter [Wed, 22 Jul 2020 13:56:22 +0000 (15:56 +0200)]
implement AsyncSeek for AsyncIndexReader
Requires updating the AsyncRead implementation to cope with byte-wise
seeks to intra-chunk positions.
Uses chunk_from_offset to get locations within chunks, but tries to
avoid it for sequential read to not reduce performance from before.
AsyncSeek needs to use the temporary seek_to_pos to avoid changing the
position in case an invalid seek is given and it needs to error in
poll_complete.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Thomas Lamprecht [Wed, 22 Jul 2020 14:01:50 +0000 (16:01 +0200)]
backup: list images: handle walkdir error, catch "lost+found"
We support using an ext4 mountpoint directly as datastore and even do
so ourself when creating one through the disk manage code.
Such ext4 ountpoints have a lost+found directory which only root can
traverse into. As the GC list images is done as backup:backup user
walkdir gets an error.
We cannot ignore just all permission errors, as they could lead to
missing some backup indexes and thus possibly sweeping more chunks
than desired. While *normally* that should not happen through our
stack, we had already user report that they do rsyncs to move a
datastore from old to new server and got the permission wrong.
So for now be still very strict, only allow a "lost+found" directory
as immediate child of the datastore base directory, nothing else.
If deemed safe, this can always be made less strict. Possibly by
filtering the known backup-types on the highest level first.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Tue, 21 Jul 2020 17:32:25 +0000 (19:32 +0200)]
ui: add calendar event selector
modelled after the PVE one, but we are not 1:1 compatible and need
deleteEmpty support. For now let's just have some duplicate code, but
we should try to move this to widget toolkit ASAP.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Thomas Lamprecht [Tue, 21 Jul 2020 11:24:48 +0000 (13:24 +0200)]
parse_calendar_event: support the weekly special expression
While we do not yet support the date specs for CalendarEvent the left
out "weekly" special expression[0] dies not requires that support.
It is specified to be equivalent with `Mon *-*-* 00:00:00` [0] and
this can be implemented with the weekday and time support we already
have.
Adding a note about the garbage collection's grace period due to the
default atime behavior should help to avoid confusion as to why space is
not freed immediately.
certain filesystems cannot be used as chunkstores, because they don't
support 2^16 subdirectories (e.g. ext4 with certain features disabled
or ext3 - see ext4(5))
reported via our community forum:
https://forum.proxmox.com/threads/emlink-too-many-links.73108/
manifest: revert canonicalization to old behaviour
JSON keys MUST be quoted. this is a one-time break in signature
validation for backups created with the broken canonicalization code.
QEMU backups are not affected, as libproxmox-backup-qemu never linked
the broken versions.
Thomas Lamprecht [Wed, 15 Jul 2020 06:33:22 +0000 (08:33 +0200)]
cached user info: check_privs: print privilege path in error message
As else this is really user unfriendly, and it not printing it has no
advantage. If one doesn't wants to leak resource existence they just
need to *always* check permissions before checking if the requested
resource exists, if that's not done one can leak information also
without getting the path returned (as the system will either print
"resource doesn't exists" or "no permissions" respectively)
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
ui: running task: increase active limit we show in badge to 99
Two digits fit nicely, and the extra plus for the >99 case doesn't
takes that much space either. So that and the fact that 9 is just
really low makes me bump this to 99 as cut-off value.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
the idea is to blend in when no task is running, thus no
background-color there. When tasks are running use the proxmox
branding guideline dark-grey, it isn't used as often so it should
fall into ones eye when changing but it has some use so it doesn't
seems out of place.
Reduce the border radius by a lot, so that it seems similar to the
one our ExtJS theme uses for the buttons outside - the original
border radius seems like it comes from the time where this was
intended to be a floating badge, there it'd make sense but as
integrated button one this seems to fit the style much more.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
When creating a new datastore the basedir is only owned by the backup
user if it did not exist beforehand (create_path chowns only if it
creates the directory), and returns false if it did not create the
directory).
This improves the experience when adding a new datastore on a fresh
disk or existing directory (not owned by backup) - backups/pulls can
be run instead of terminating with EPERM.
Tested on my local testinstall with a new disk, and a existing directory:
it does not make sense to check if the worker is running if we already
have an endtime and state
our 'worker_is_active_local' heuristic returns true for non
process-local tasks, so we got 'running' for all tasks that were not
started by 'our' pid and were still running