The IPv6 provisioning support has no page in the admin guide, so two
requirements that a cluster cannot work without are only visible in the code.
Only netboot=grub2, grub2-http and grub2-tftp get a bootfile-url in a DHCPv6
reservation, and the grub2 image under /tftpboot/boot/grub2 is not signed, so
firmware that enforces Secure Boot downloads it and refuses to start it.
The page states both requirements, how xCAT predicts the DHCPv6 client
identifier from vpd.uuid, what the configdhcpv6duid postscript does, and the
static address that nodeset puts on the kernel command line.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The SELinux page carried the repair in two blocks. The first mapped the
published artifacts and ran restorecon over the image root. The second mapped
the publication directory and excluded both guest trees. An operator with
site.installdir=/srv/xcat-install who reads the page in order therefore
relabels before any mapping covers that tree, and rootimg/etc/shadow resolves to
var_t. The exclusion added afterwards keeps the var_t it finds, because it only
stops the next relabel from changing the label again.
The order also cost the artifacts. file_contexts keeps the expressions in the
order they are read and the last one that matches a path wins, so the exclusions
of the second block beat the artifact mappings of the first. Where the image
root carries a rootimg component in the middle, the exclusion matches the
artifacts as well, and the five expressions with a wildcard, the initrds and the
packed images, resolve <<none>>. kernel and hypervisor carry no wildcard and win
either way, which is why the defect showed on a subset.
The page is now one sequence: the publication directory, both guest exclusions,
the artifact mappings, then one restorecon. It says what each step orders and
why.
xCAT-test/native/selinux_image_repair.t passes 381 of 381 checks on this
sequence and fails 47 on the order of the parent commit. selinux_docs.t fails 3
of 30 on the parent commit.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The SELinux page told the operator to run chcon -R -t public_content_t over the
image root to repair a published artifact that kept an older type. That command
walks into rootimg and rootimg-statelite, so it replaces the labels of the guest
operating system with public_content_t. The xcat.fc comment carried the same
advice. chcon also writes a type that no mapping selects, so the next relabel of
a path that has a mapping reverts it: the repair held only until then.
The page now maps each published artifact with semanage fcontext and applies the
mapping with restorecon. Each expression names one regular file directly below
the image root, so no path in either guest tree matches one, and both trees keep
the <<none>> of the packaged mappings. The procedure reads the publication
directory from site.installdir, and a site that moved it maps the configured
directory and keeps both guest trees out of that mapping. The page says why a
tree mapping over the image root is the same defect as the chcon.
xCAT-test/native/selinux_image_repair.t passes 170 of 170 checks with the
documented procedure. With the chcon of the previous page it fails 101, 36 of
them guest labels. selinux_docs.t fails 7 of 26 on the parent commit.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The file contexts leave the published kernel of a middle-rootimg image
root out of the relabel. The comment and the SELinux page read that as
the kernel being unlabeled, httpd refusing to serve it, and that osimage
losing netboot, with restorecon as the repair. Three of those four
claims are wrong.
Measured. An entry of <<none>> keeps the label a path has: restorecon
-R, restorecon -RF and setfiles -F all left a kernel at default_t, while
the same setfiles call relabeled a path that has a rule. A new file
inherits the type of its directory, so a kernel that genimage writes
into an image root already made public_content_t is public_content_t. In
the policy with the xcat module loaded, httpd_t holds one allow rule to
read public_content_t and none for default_t. So the ordinary case
serves and netboot works, and restorecon is not the repair for the case
that does not: chcon -R -t public_content_t over the image root is, and
a later relabel leaves that value alone.
selinux_file_contexts.t now starts half the custom roots at
public_content_t, which is what genimage produces, and half at
default_t, and requires each to keep its own type. It fails 324 of its
618 checks against the previous xcat.fc.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
A file context of <<none>> stops restorecon changing the label. It does
not clear it. A tree that genimage creates under an image root that is
already public_content_t inherits that type, measured on an enforcing
el10 host: mkdir -p rootimg/common/sub under a public_content_t parent
gives every new component public_content_t, and the same tree under a
default_t parent gives default_t.
Say what the label is, and name the case that costs something: a tree
older than the package keeps default_t, and restorecon -RF over the
image root sets it.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The security page said xcatd runs in unconfined_service_t and that
xCAT-selinux defines no type. The module defines xcat_t for xcatd and
the client commands, and xcat_image_t for the image build, and declares
both permissive. The older page placed a service node and a compute node
outside the scope because xCAT does not select their mode, and the branch
ships noderes.selinux and site.selinux.
Both pages now state what the module does: the two permissive domains,
the transitions to tftpd_t and chronyd_t for the services xCAT starts,
the mode a node resolves, and the goal of an enforcing management node.
The scope page also records that a published common or overlay directory
of an image root called rootimg is unlabeled, for the same reason as a
published kernel there.
selinux_docs.t fails 6 of its 18 checks without this change, on rome01.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
%post writes /etc/xcat/xcatroot.env and %preun removes it, so the file
is absent on rpm --justdb, before %post runs, and after an interrupted
install. Both subtests of xcatd_relocated_startup.t have the file
present, so neither reads what happens without it.
Add a subtest that removes the file: each EnvironmentFile of the unit
carries a leading -, which makes a missing file not an error, XCATROOT
is the default of the unit, and the daemon starts with that default
instead of the installed prefix. Dropping the default XCATROOT from the
unit fails checks 4 and 7 of that subtest.
Record the trade the file contexts make for a rootimg component in the
middle of an image root: the guest labels are kept, which leaves the
published kernel of that one layout unlabeled, so httpd cannot serve it
and that osimage cannot netboot. The note names the recovery.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The page states that the context belongs to the file in /opt/xcat. The
prefix is now recorded at install time, so a relocated package starts
the daemon of its own tree, and the file contexts still name /opt/xcat.
Say both, so a reader of a relocated installation knows the entrypoints
need their own entries.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The branch changed what a cluster runs and added no documentation, while
the quick start guide still told the operator to write SELINUX=disabled
on the management node before installing xCAT.
docs/source/advanced/security/selinux.rst says which machine gets which
mode, why a stateless node runs with SELinux disabled, how to set
noderes.selinux and site.selinux, and that xcatd, in.tftpd, kea-dhcp4
and goconserver run in unconfined_service_t, so the work is about
running with SELinux enforcing and not about confining xCAT.
It also covers the upgrade: recordselinuxdefault runs on INITIALINSTALL
only, so an upgraded cluster has no site.selinux value, every node keeps
being provisioned with SELinux off, and reprovisioning is what gives a
node the new mode.
The quick start step now says SELinux can stay enabled and links there.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The page was written for a single domain and three entrypoints. It now names
both domains, the five client entrypoints including xcatDBcmds and the
genimage wrapper, and the two per-OS build programs.
It also states what each domain holds, so a reader does not take the split
for more than it is: xcat_image_t may relabel every type including the
security ones, and xcat_t keeps manage and relabel of non-security files
because packimage and liteimg run inside xcatd. And it records that a
permissive declaration does not cover a transition into a stock domain.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The pull request adds a process domain and says nowhere which machine it
covers, which commands enter it, or why the domain is permissive while the
host is enforcing.
The new page names the scope as the management node, lists the three
entrypoint files and the rule that the unit must exec xcatd and not a shell,
and states that the permissive declaration is in the policy source and a
later release removes it. It gives the commands that read the domain of
xcatd and the enforced denials of a run, with the --input-logs finding:
ausearch read the journal on an EL10 node whose log held 1032 denials.
It also records why a confined user domain gets no transition, and that the
file permissions still decide what a user can read.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
The bats stub returned one exit code for every "chdef netboot=", and the
helper calls chdef twice: once to set the method to ipxe, and once to put
the saved method back. With the stub failing both, "a netboot method that
cannot be put back fails the case" passed on the FORWARD failure, so
removing restore_rc=1 from the restoring branch left all 18 tests green.
The test named a defect it could not detect.
The stub now counts its netboot calls and fails the one the test names,
so the forward chdef succeeds. The test asserts that forward call is in
the call log, and matches the message only the restoring branch prints.
Removing restore_rc=1 now fails test 8 and no other.
docs/source/advanced/ipxe/index.rst said both methods run the same boot
scripts, which stopped being true when the shim line became conditional
on the method. It now says what differs, and states that Secure Boot
reaches netboot=ipxe only: xnba.efi is unsigned, and grub2 nodes are out
of scope.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
This PR adds a shim to the UEFI boot scripts of netboot=ipxe nodes, so a node
with Secure Boot on reaches its kernel.
The firmware signature database does not hold the keys of a Linux
distribution. A UEFI node with netboot=ipxe loads snponly-shim.efi and
snponly.efi, both signed, and then stops at the kernel with "Verification
failed: Security Policy Violation".
xCAT::SecureBoot::shim_url_path names a shim for one node. The shim of the
install source comes first, because its vendor certificate signs the kernel of
its own distribution. copycds keeps the names of the DVD, so the module tries
EFI/BOOT/BOOTX64.EFI, EFI/BOOT/bootx64.efi and EFI/boot/bootx64.efi: EL and
openEuler ship the first, openSUSE the second and Ubuntu the third. When no
install source carries a shim, the module names the shim of ipxe-xcat, which
verifies a kernel that a MOK key signs.
xnba.pm writes the shim line after imgload in the UEFI script of an ipxe node,
and mknb.pm writes it in the UEFI discovery scripts. iPXE fetches the shim only
when it cannot load the kernel itself, so a node with Secure Boot off reads no
shim. netboot=xnba is unchanged: no key signs xnba.efi.
The elilo and ESXi paths and the Genesis kernel stay unsigned, and the iPXE
documentation now says so. The validation matrix gets a row for a UEFI install,
a netboot image and discovery with Secure Boot on.
secureboot_ipxe_node_shim.t and secureboot_genesis_shim.t drive nodeset and
mknb into a scratch TFTP root. Four assertions of ten and two of seven fail
without this change.
Signed-off-by: Daniel Hilst <392820+dhilst@users.noreply.github.com>
A new page describes the ipxe and xnba netboot methods: the files of the
upstream loader, UEFI Secure Boot through the shim and its limits, the
client test, the loader of nodes without a definition, the move of
nodes to the ipxe method and back, and the known limits of the upstream
loader.
The node guide, the noderes, node, group and nodeset pages list the ipxe
method, and the value lists mark xnba deprecated. The examples that
define an x86 node use netboot=ipxe. The mixed cluster and uninstall
guides name ipxe-xcat, and the DHCP validation matrix gets the iPXE
loader rows, Secure Boot included.
The guides said riscv64 covered EL10 only, and the riscv64 page listed
Ubuntu as unsupported. Both support matrices now carry the architecture for
Ubuntu, and the riscv64 page describes the Ubuntu paths: the loader copycd
builds from the media, the installer needing none of the accommodations
EL10 requires, the ports archive the packages come from, and a management
node running on riscv64. The 26.04 media need the RVA23 profile, which is
recorded as a limitation.
Signed-off-by: Vinícius Ferrão <2031761+viniciusferrao@users.noreply.github.com>
Add the riscv64 page to the cluster management guide (UEFI + grub2 boot
path, discovery through mknb's grub2 network configurations, stateful and
stateless provisioning, the dependency picture for a management node on
riscv64, limitations), list riscv64 in the node object attributes and in
the support matrices, add the architecture to the cross-build page for
stateless images, extend the grub2 install guide (and fix its swapped
x86_64/aarch64 file names), the uninstall package lists, the DHCP backend
validation matrix and the mknb/genimage man pages, and carry the riscv64
schema values into the generated nodetype, osimage, noderes, node and
group references.
xcat.conf was installed as an ordinary payload file and then deleted and
recreated from the Apache-version template in %post. rpm therefore held
no record of what was on disk, and an upgrade replaced an edited file
silently, leaving neither .rpmnew nor .rpmsave. A site that had added
Indexes to the /install block lost it on upgrade and directory listings
began returning 403.
Select the Apache 2.2 or 2.4 configuration at build time, using the same
distribution macros the rest of the spec already relies on, and mark
both /etc/httpd/conf.d/xcat.conf and /etc/apache2/conf.d/xcat.conf as
%config(noreplace). rpm then keeps a modified file and installs the new
vendor version alongside it as xcat.conf.rpmnew.
The old payload recorded the 2.2 file while %post wrote the 2.4 one, so
rpm cannot distinguish a stock file from an edited one across the
transition. A migration compares the active file with the templates the
outgoing package saved under conf.orig and removes it only when it is a
regular file still byte-for-byte identical to one of them. A stock
upgrade then completes without an unnecessary .rpmnew, and anything that
differs is left untouched.
That migration runs in %pretrans, not %pre. rpm fixes each config file's
fate before %pre, so removing the active file there can happen after rpm
has already resolved to write only xcat.conf.rpmnew, leaving the system
with no active configuration at all. %pretrans runs before that decision.
It is an embedded Lua scriptlet because a pre-transaction scriptlet
cannot rely on any dependency being unpacked yet, which also means the
comparison needs no external tool.
bc was needed only by the version check the service-node package no
longer performs.
The Apache directives are unchanged. Document a later-loading conf.d
file as the place for site rules, since that survives upgrades without a
merge.
genpdu read negative-capable sensors from the unsigned value column,
which the MIB leaves undefined for them: a PX4 answers it with 0 for
inlet reactive power and carries the reading in the signed column. Use
SensorSignedMinimum to pick the column, as the MIB specifies. The
unsigned column is still needed for active energy, whose range exceeds
Integer32 and reads 0 in the signed one.
An unreachable PDU or a wrong credential looked like a PDU without
switched outlets, since the switching probe was the first exchange with
the device. Probe pduCount first (implemented by every PDU2 agent,
verified on PX2, PX3 and PX4) and return no session when it fails, so
callers print "Couldn't connect to <pdu>". Warn when pduCount is not 1
rather than refusing: a linked primary still answers for its own
outlets.
Add unit tests for the genpdu sensor, session and inventory paths.
Signed-off-by: Kilian Cavalotti <kilian@stanford.edu>
Add a new pdutype (genpdu) for PDUs implementing the Raritan PDU2-MIB.
A single MIB covers the Raritan PX2/PX3/PX4/PXC/SRC/PXO/BCM series, the
Server Technology PRO3X/PRO4X series, the Legrand intelligent PDUs, and
all PDUs following the PDU2-MIB.
Supports rpower (whole-PDU and per-outlet), rinv and rvitals over SNMP
v1, v2c or v3, with credentials read from the pdu table).
Unlike the existing types, sensor units and decimal precision are read
from the MIB per sensor rather than hardcoded, so readings are correct
across models that report differing precision for the same sensor.
Outlet switching capability is probed at connect time, so metered-only
models report a single "unsupported" message instead of a per-outlet error.
rspconfig is not supported for genpdu; PDU linking and BCM2/PMC power
meters are out of scope.
Tested on Raritan PX4-5851-E7V2 (fw 4.2.10.5-50400, switched),
PX3-1901U-N1 and PX3-1901U-N1A6 (fw 4.0.20.5-49038, metered), and
PX2-1901U-N1A6 (fw 4.0.20.5-49038, metered).
Signed-off-by: Kilian Cavalotti <kilian@stanford.edu>
Add a shared OMAPI policy helper for ISC DHCP and DDNS so administrators can select the key name, signing algorithm, and omshell path from the site table while preserving the existing xcat_key hmac-md5 default.
Keep local ISC updates from hanging indefinitely when omshell does not exit, and use a static host-declaration fallback for local Ubuntu ISC releases where omshell is unstable for xCAT host updates.
Co-authored-by: gskouson <1507929+gskouson@users.noreply.github.com>
The CUDA docs were frozen at CUDA 9.2 / RHEL 7.5 / Ubuntu 14.04 since
2019. Update to cover all currently supported OS and architecture
combinations (EL 7-10, Ubuntu 20.04-24.04, x86_64/ppc64le/sbsa).
Consolidate the version-specific repo and osimage pages into generic
guides that use placeholder variables, reducing 7 files to 2 while
covering more OS versions. Both online (direct NVIDIA repo URL) and
offline (dnf download / apt download mirroring) workflows are
documented.
All NVIDIA repository URLs validated against
developer.download.nvidia.com/compute/cuda/repos/ and confirmed
accessible with valid repodata.
Addresses #7373
The default xCAT Apache configuration shipped with Options Indexes
enabled for the /install and /tftpboot directories. This allowed
unauthenticated users to browse directory listings, disclosing the
full tree of postscripts, boot files, and (in production deployments)
potentially kickstart files with password hashes, custom scripts with
embedded credentials, and cluster topology details.
Replace Options Indexes with -Indexes in all four shipped Apache config
files (MN and SN, Apache 2.2 and 2.4 variants). Direct file access
by known path continues to work, so all provisioning workflows are
unaffected. Directory browsing for /xcat-doc is preserved as it
contains only public documentation.
Additionally, add an Apache hardening guide documenting recommended
permissions for sensitive directories under /install, network binding
best practices, and IP-based access control options.
Addresses #7450
Docker container lifecycle management (mgt=docker, mkdocker, rmdocker,
lsdocker) was added in 2015-2016 as an experiment targeting Docker API
v1.22 on Ubuntu only. Documentation and man pages were deliberately
removed in 2019 (PRs #6222 and #6324) with the original developer's
approval, noting that "the interface of Docker has become very simple
right now, so there is no value for xCAT to offer such functions."
The plugin was still being shipped but has had no functional code changes
since April 2016, was never listed as a valid mgt value in Schema.pm,
and no user ever filed an issue about it.
Removed:
- xCAT-server/lib/xcat/plugins/docker.pm (1,142 lines)
- xCAT/postscripts/setupdockerhost
- xCAT-server/share/xcat/scripts/setup-dockerhost-cert.sh
- xCAT-test/autotest/testcase/dockercommand/ (test cases)
- Docker attribute definitions in Schema.pm
- Client symlinks (mkdocker, rmdocker, lsdocker)
- Usage entries and dockerhost cert handling in credentials.pm
- Docker attribute documentation in man7 pages
The "Running xCAT in Docker" documentation (dockerized_xcat/) is
retained as it documents containerizing xCAT itself, not the removed
mgt=docker feature.
Closes#7518