| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
|
|
| |
Raise error when _connect() returns non zero status.
Signed-off-by: Trevor Benson <trevor.benson@gmail.com>
|
|
|
|
| |
Signed-off-by: Trevor Benson <trevor.benson@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
| |
'get-mempool-info' without specifying a netdev will capture
all the dpdk mempools that exist including ones no longer
associated with a netdev but yet to be freed.
'driverctl list-persisted' will list the overrides that are
persistent through reboots.
Signed-off-by: Kevin Traynor <ktraynor@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Collect almost all params, except those explicitly excluded.
Those excluded are known to be less useful debugging, and/or
known to be very large.
Some params are still split out to help limit size.
Some very large (but possibly useful) param are relegated to `--all-logs`
Signed-off-by: Nathaniel Clark <nclark@whamcloud.com>
|
|
|
|
|
| |
Closes: #3439
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
|
| |
Expand the cockpit plugin to capture configuration
and output of cockpit-bridge. Also remove deprecated
command 'remotectl'.
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
| |
Also add `opa-fm` service check
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently, the arcconf getconfig and GETLOGS commands are added with the
assumption that the system has only one controller with ID 1. It is
possible that a system can have multiple controllers, so commands need
to be added to extract information for all available controllers.
The arcconf utility does not have a dedicated command to get the list of
controllers available in the system. Therefore, parse the 'arcconf list'
output to obtain the controller ID for all controllers in the system,
and add getconfig and GETLOGS commands for each available controllers.
Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com>
Suggested-by: Borislav Stoymirski <borislav.stoymirski@bg.ibm.com>
|
|
|
|
| |
Signed-off-by: Adam Vest <adam.vest@canonical.com>
|
|
|
|
|
|
|
|
| |
ARM devices often don't have SMBIOS information, so dmidecode does not
return anything useful. Add the device tree model information and
compatible string, which at least provides the basics.
Signed-off-by: Simon Glass <sjg@chromium.org>
|
|
|
|
|
|
|
|
|
| |
Having nvidia-persistenced service status helps debugging issues
with NVIDIA kernel driver behaviour with persistence mode on.
Signed-off-by: Sourabh Jain <sourabhjain@linux.ibm.com>
suggested-by: Borislav Stoymirski <borislav.stoymirski@bg.ibm.com>
Tested-by: Borislav Stoymirski <borislav.stoymirski@bg.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch reports the output of the following Mellanox firmware
commands:
mlxdump
mstconfig
mstdump
mstflint
mlxreg
mlxlink
mlxcables
mst
Additionally, if the user allows system changes, we will atempt
to start the mst device, add the cables and stop the device in
the end.
Signed-off-by: Alin-Gabriel Serdean <aserdean@ovn.org>
|
|
|
|
|
|
|
|
|
|
|
| |
Commit ac17ca4f4 removed the command build_sphinx from
setup.py, but the README.md file still references it.
This patch offers an alternative for uses wishing to generate
html docs, using the command sphinx-build.
Related: RH: RHEL-17924
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Collect current auth profile details.
Resolves: #3438
Relevant: RHEL-18576
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
| |
As part of the discovery is also discovery-toolbox plugin
we should collect also logs from this plugin which may
be helpful in debugging issues.
Signed-off-by: Jan Jansky <jjansky@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Collect up to 14 days of dynflow data instead of one month.
Also, change granularity of the plugin option from months to days - we
are almost never interested in tasks older than one month.
Resolves: #3437
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
As SoSComponent already works with --tmp-dir and it can be declared in a
preset, we need to move whole options evaluation from SoSReport to
SoSComponent constructor.
Resolves: #3425
Closes: #3432
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When running sphinx-build to generate html documentation,
we get the following warning:
sos/report/plugins/__init__.py:docstring of
sos.report.plugins.Plugin.collection_file:15:
WARNING: Inline literal start-string without end-string.
Related: RH: RHEL-17923
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
|
| |
These modules are not present anymore, at least on RHEL8.
The modules are not needed to execute 'vmcp' commands.
Signed-off-by: Renaud Métrich <rmetrich@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When adding or deleting preset (and also when detecting no plugin leaves
enabled), sos.report.execute must cleanup its working dir.
Since we don't use return value of the execute method, we can skip the
return command.
Resolves: #3430
Closes: #3431
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
| |
Capture PV headers and metadata when a new option,
lvm2.metadata , disabled by default, is specified
as plugin option.
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
| |
Clean up one old reference to an obsolete option.
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
EDPM (External DataPlane Management) will be the new way to deploy
compute nodes startin from RHOSP 18 which will replace TripleO/Director
- Added openstack_edpm plugin
- Added custom path for config and logs in
frr/libvirt/logrotate/openstack_neutron/openstack_nova/ovn_host
modules
- Tagged libvirt/openstack_neutron/openstack_nova/ovn_host modules
as `openstack_edpm`
Signed-off-by: Roberto Alfieri <ralfieri@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Forbid collecting some files and dirs from /etc/ansible that are not
interesting for any investigation but might potentially collect
sensitive data.
Resolves: #3423
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* python3-setuptools is still required for building the package
* python3-yaml is required for build for tests with the
conformance_tests
* Add the same ignores as dowstream build, so that we don't have to
keep changing these. Once the avocado is packaged, we can take this
out
* Update image for the deb creation, so that it can be minimal
* Fix lxd runtime, so that if the root device is not available, it
will not look at the storage pools
Downstream build machines are much more minimal compared to gce
instances, so we require more dependencies to be added.
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
|
| |
Closes: #3278
Co-authored-by: Pravesh Gaire <pravesh.gaire@canonical.com>
Signed-off-by: Pravesh Gaire <pravesh.gaire@canonical.com>
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
| |
Collect some additional configuration files based on the description of
man 5 NetworkManager.conf
Signed-off-by: Hiroyuki Yasuhara <h.yasuhara@fujitsu.com>
|
|
|
|
|
| |
Resolves: SET-383
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
|
| |
Livepatch on-prem server support
Resolves: SET-382
Signed-off-by: Adam Vest <adam.vest@canonical.com>
|
|
|
|
|
|
|
| |
Related: #3398
Related: #3383
Resolves: SET-326
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
New version of juju uses colorisation, and therefore juju status
and json.loads doesn't load the juju status correctly. By using
--no-color based on the version of juju this should fix this
particular use-case
Resolves: #3399
Resolves: SET-339
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
| |
Remove file /var/snap/maas/current/http/certs/regiond-proxy-key.pem so
that it is not collected by sosreport
Signed-off-by: Jorge Merlino <jorge.merlino@canonical.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Initial start of building and testing the deb package
This will improve the overall testing of sos over the Ubuntu releases,
and also helps with the basis of thenadding further stagetwo testing in
the future. This will help to pick up issues automatically rather than
the manual testing.
Add ubuntu-latest for deb CI build and testing. Update the GCE images
for all ubuntu images for CI testing
Co-authored-by: Nikhil Kshirsagar <nikhil.kshirsagar@canonical.com>
Signed-off-by: Nikhil Kshirsagar <nikhil.kshirsagar@canonical.com>
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
|
|
|
|
| |
Due to the deprecation of the existing "pcs property list",
this patch checks the version of pcs and executes
"pcs property config" as an alternative."
Signed-off-by: Yuki Inoguchi <inoguchi.yuki@fujitsu.com>
|
|
|
|
|
|
|
|
|
| |
Gather details about branch, so it complements
the version number (which can be empty sometimes).
Fixes: #3404
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
| |
Resolves: #3408
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The `get_pty` parameter for remote executed commands was both a bit of a
misnomer and applied too broadly.
Refactor this to `use_shell` to be more obvious about what the intent
behind the option is, and default all transports to `False`, so that by
default we do not wrap any commands in a bash shell.
This may be overriden on a per-transport basis via the ned `_need_shell`
property within transport subclasses. Further, this facility has been
expanded to be allowed on a per-command basis from
`SoSNode.run_command()` and wherever that is linked.
Related: #3399
Related: #3400
Signed-off-by: Jake Hunsaker <jacob.r.hunsaker@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
| |
In newer versions of tuna, the option -P has been
substituted for 'show_threads'.
Test if the new option is present in 'tuna --help'
and if so, use it.
Closes: RH SUPDEV-150
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
|
| |
Sometimes the update_sources attribute is not available so cannot
build the list
Signed-off-by: Arif Ali <arif.ali@canonical.com>
|
|
|
|
| |
Signed-off-by: Marek Czernek <marek.czernek@suse.com>
|
|
|
|
|
|
|
|
| |
This is a followup for 56904018045fdbdad666e55f0a850bc365809e50
I hereby declare that you can do whatever you want with this commit:
Signed-off-by: Miro Hrončok <miro@hroncok.cz>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This commit removes the duplicate "rpcinfo_-p_localhost" log
from the sos_commands/nfs directory. Currently, the same log
file is being createdin sos_commands/sunrpc/ directory,
as shown below:
# ls sos_commands/nfs/rpcinfo_-p_localhost
sos_commands/nfs/rpcinfo_-p_localhost
# ls sos_commands/sunrpc/rpcinfo_-p_localhost
sos_commands/sunrpc/rpcinfo_-p_localhost
Since rpcinfo command is included in rpcbind package,
let's remove the one in the nfs directory.
Signed-off-by: Yuki Inoguchi <inoguchi.yuki@fujitsu.com>
|
|
|
|
|
|
|
| |
Capture two ImportError exceptions when
calling __import__()
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
|
|
| |
Resolves: #3393
Signed-off-by: Nikhil Kshirsagar <nikhil.kshirsagar@canonical.com>
|
|
|
|
| |
Signed-off-by: Arnaud Veron <arnaud.veron@opensvc.com>
|
|
|
|
|
|
|
| |
Resolves: #3391
Closes: #3390
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
driverctl commands give info about which drivers are used
for devices. They list PCI address and driver used e.g.
0000:01:00.0 vfio-pci
This can usually be correlated to the device type through lspci.
However, lspci may not be available if only some plugins were
captured as part of the report.
Even when lspci info is available, it is still convienent to have
the device type co-located with the driver and override information.
Add verbose option to driverctl commands that support it to also
list device type. e.g. (wrapped for line length)
0000:01:00.0 vfio-pci
(82599ES 10-Gigabit SFI/SFP+ Network Connection
(Ethernet 10G 4P X520/I350 rNDC))
Signed-off-by: Kevin Traynor <ktraynor@redhat.com>
|
|
|
|
|
|
|
|
| |
Likewise in #3379, scrub the password also in pulpcore plugin.
Resolves: #3389
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
| |
Related: RH SUPDEV-136
Signed-off-by: Jose Castillo <jcastillo@redhat.com>
|
|
|
|
| |
Signed-off-by: Pablo Acevedo Montserrat <pacevedo@redhat.com>
|