| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
| |
The rhevm plugin has a 'vdsmlogs' option to pass in a set of log
files to be collected. When run with '-a' (or with
'-krhevm.vdsmlogs) this evaluates to 'True' (a bool) and causes
an exception when this is passed to add_copy_specs().
Always treat the value as a string.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Prior versions of sos enable debug logging to the embedded log
file (sos_logs/sos.log) when a single '-v' is given. Restore this
behaviour and ensure that command-not-found messages are reported
at 'info' rather than 'warning' level.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
| |
Commit 4ab4b08 inadvertently removed the postprocessing rules for
luci configuration. Revert that part of the commit.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Since the ldap plugins add files and post-processing methods at
multiple class levels (Ldap, RedHatLdap etc.) derived classes
must explicitly call their parent class's postproc() method to
apply all substitutions to collected data.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The cluster plugin used an obsolete sos-2.2 method to determine
the command output directory. This causes an excaption at runtime
since the referenced properties no longer exist.
The crm_report script also expects a --from date and will not
collect data unless this is passed. Default to passing a value 72
hours before the current time and add a 'crm_from' option to the
cluster module to allow the user to override this.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|\
| |
| |
| |
| |
| | |
Fixes #230
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
| | |
|
| |
| |
| |
| | |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
below
sos/plugins/anacron.py:
- add one line description during verbose output for anacron
sos/archive.py, sos/utilities.py:
- make sure to return a string after a Popen call
which initially returns binary
sos/plugins/named.py:
- Remove unused import of commands
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The formatting of the template preamble used a narrow terminal
width since there is no native python method for obtaining the
actual terminal geometry and the common methods for doing this
are non-portable (e.g. Windows specific ioctls or shelling out
to stty).
This causes the top of the message to scroll off the screen on
a standard 80x25 terminal (e.g. putty or a non-resized
gnome-terminal or xterm window).
Remove the paragraph indentation and increase the assumed width
from 58 to 72 chars. This causes the message to fill a standard
80x25 terminal without scrolling.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
| |
| |
| |
| | |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|/
|
|
|
|
|
|
|
|
|
|
| |
The --upload option has a lot of limitations and has not been
widely used. Recent trends are to handle uploading data in
higher-level tools (e.g. redhat-support-tool or web based
management UIs) and the python ftp library does not support
modern requirements like HTTP proxy traversal or encryption.
Fixes Issue #217
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
|
|
|
| |
Fixes #229
Signed-off-by: Raphael Badin <raphael.badin@canonical.com>
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
|
|
|
| |
Fixes #227
Signed-off-by: Raphael Badin <raphael.badin@canonical.com>
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
|
|
|
| |
Fixes: #222
Signed-off-by: Louis Bouchard <louis.bouchard@canonical.com>
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
|
|
|
|
| |
This rename happened in now current release OpenStack Havana.
Fixes #220
Signed-off-by: Alan Pevec <apevec@redhat.com>
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
|
|
|
|
| |
The last commit adds a log message for commands that time out but
is inconsistent with other messages that include an external
command string. Quote the command for better readability.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
| |
The anaconda kickstart file is Red Hat specific and should not
be collected in the general plugin. It may also contain plaintext
or encrypted passwords - these are handled properly in the
anaconda plugin itself but are missing from the copy in general.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit 6ea48cb changed the action for the -k/--plugin-option
command line argument from 'extend' to 'append'. This breaks the
handling of options that are chained with a ',':
logs.logsize 1,logs.all_logs max size (MiB) to collect per syslog file
logs.all_logs off collect all log files defined in syslog.conf
vs:
logs.logsize 1 max size (MiB) to collect per syslog file
logs.all_logs on collect all log files defined in syslog.conf
Revert to an 'append' action for this switch.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Several log messages used the 'error' level when reporting an
event that is not necessarily an error (and does not prevent sos
from continuing).
Reclassify these to the 'warning' log level instead.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Commit 6ea48cb changed the default log level for console output
from WARNING to ERROR and reclassified several messages.
This causes plug-in validation messages to be silent at the
default verbosity.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The networking module parses the link information in the output
of the 'ip -o addr' command to determine a list of ethernet
interfaces on which to run ethtool.
An upstream change in the iproute package appears to have
introduced a regression in the output of this command which causes
link information to not be displayed:
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000\ link/ether 52:54:00:ce:9d:6d brd ff:ff:ff:ff:ff:ff
2: eth0 inet 192.168.122.31/24 brd 192.168.122.255 scope global eth0\ valid_lft forever preferred_lft forever
2: eth0 inet6 fe80::5054:ff:fece:9d6d/64 scope link \ valid_lft forever preferred_lft forever
vs:
2: eth0 inet 192.168.122.31/24 brd 192.168.122.255 scope global eth0\ valid_lft forever preferred_lft forever
2: eth0 inet6 fe80::5054:ff:fece:9d6d/64 scope link \ valid_lft forever preferred_lft forever
This breaks detection of ethernet interfaces and all ethtool data
collection.
This was introduced in the following iproute commit:
commit af9d406f99853848054162ed7aefbe71dc03e433
Author: Mike Frysinger <vapier@gentoo.org>
Date: Mon Aug 13 08:09:52 2012 -0700
Fix regression with 'ip address show'
Which appears to fix a different problem (i.e. the change to link
info was unintended).
Make this more robust by switching the networking module to use
'ip -o link' instead which explicitly requests the required
information.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The changes to reinstate the --build option introduced two
regressions:
- Running sosreport with no otions generates a backtrace due to
a non-existant _build member in the SoSOptions object
- Running sosreport without --build leaves the temporary archive
directory in places with the sos_logs directory containing
sos.log and ui.log due to changes in the order of log shutdown
and archive creation.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Using the print statement as though it was a function leads to
incorrect behaviour in a python2 runtime. E.g.:
print()
print( "Some examples:")
print()
Produces:
()
Some examples:
()
Instead replace use of the print keyword with a call to the six
module's print_() function.
Fixes Issue #219
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
The urllib import via six.moves in the plugins module is unused.
This causes an exception on some distributions when attempting to
import the moved module.
Drop the import since nothing in the plugin classes is using it.
Fixes Issue #218
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
| |
The 'all_logs' path of the logs plugin has three leftover debug
print statements. Remove them to avoid cluttering the console
output when the option is used.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
| |
openstack-* is longer than the allotted 15 spaces, increased to 20 to make
things line up again.
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|
|
|
| |
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
|\
| |
| |
| |
| |
| | |
Forward merges from master to make this PR mergeable again.
Signed-off-by: Adam Stokes <adam.stokes@ubuntu.com>
|
| |
| |
| |
| |
| |
| |
| | |
The PKI files that should be omitted are in /etc/pki/entitlement,
not /etc/pki/entitlements as the plug-in currently uses.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
There has been an nfsserver plug-in for some time however no
corresponding client-side data is collected.
This was added to sos-2.2 earlier this year. This commit is a
forward port of those changes:
commit 71066b4a7d4fd97692b50bf231ad6810df5ec3d0
Author: root <root@rhel6-vm2.(none)>
Date: Wed Jul 10 17:11:52 2013 +0100
Collect additional nfs-client relevant files
Additional NFS client-side data collection:
"/etc/nfsmount.conf",
"/proc/fs/nfsfs/servers",
"/proc/fs/nfsfs/volumes"
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
| |
| |
| |
| | |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Forward port of the following sos-2.2 fix:
commit d4e3bcc3ed247d8d9a1228c3661b6456f534daf4
Author: Bryn M. Reeves <bmr@redhat.com>
Date: Mon Aug 12 18:42:48 2013 +0100
Fix traceback when sar module is force-enabled
The sar plug-in assumes that when it runs the /var/lib/sa
directory and binaries exist (they are part of checkenabled() for
this module).
This leads to a traceback when the plug-in lists a non-existent
directory on hosts where sar is not installed but '-o sar' is
specified on the command line.
Catch the exception and exit gracefully.
(Edited by bmr to improve error message logged).
Signed-off-by: David Kutálek <dkutalek@redhat.com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The foreman and katello plug-ins committed in 4650d3f use the
sos-2.2 APIs and class names. This breaks sos since the plug-ins
will not load under 3.0.
Fix up the modules to use the new names and API conventions and
ensure that they load correctly.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Releases of sos prior to 3.0 included the '--build' option to
disable creation of a compressed archive and to leave the
temporary build tree in place. This was removed with the
reorganisation of the archive classes to support in-line
archiving. Since all supported policies are now using an archive
that derives from FileCacheArchive (commit 0178d5f) the option
can be re-introduced.
Archive classes that do not accumulate files in a temporary
directory will return the path to the in-progress archive file.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The archive and temporary file handling classes expect to operate
on absolute paths. If a user specifies a relative path via the
command line --tmp-dir switch it must be normalized before being
passed to these classes to provide correct behaviour. Failing to
do this causes a variety of errors: the final archive cannot be
created since the path is constructed incorrectly and temporary
files are not cleaned up since they are not at the "expected"
location.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|