| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
|
|
| |
instead of add_copy_spec_limit()
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
|
|
|
|
| |
The difference between add_copy_spec() has been removed.
This is in preparation of removing the add_copy_spec_limit()
special case altogether.
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
|
|
| |
This is in preparation of converging to one add_copy_spec() function
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
|
|
|
| |
- a test with a single file without limit
- a test with a single glob without limit
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
|
|
|
|
| |
Originally,f there were other files matching the glob of /tmp/tmp*
then the test would fail. By using a tmpdir, we make sure that we
know what files to expect.
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
| |
Signed-off-by: Michael Adam <obnox@samba.org>
|
|
|
|
|
|
|
|
|
|
| |
add_cmd_output now gets only called if the needed ENV variables
are set to be able to connect to an OpenStack env.
Resolves: #956.
Signed-off-by: Martin Schuppert <mschuppe@redhat.com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There is a problem with six's handling of strings that end with a
backslash:
https://bitbucket.org/gutworth/six/issues/60/sixu-mishandles-string-ending-in-backslash
This leads to an ugly exception during reporting if any filename
ends with this character:
Traceback (most recent call last):
File "/usr/sbin/sosreport", line 25, in <module>
main(sys.argv[1:])
File "/usr/lib/python2.7/site-packages/sos/sosreport.py", line 1632, in main
sos.execute()
File "/usr/lib/python2.7/site-packages/sos/sosreport.py", line 1607, in execute
self.plain_report()
File "/usr/lib/python2.7/site-packages/sos/sosreport.py", line 1358, in plain_report
output = PlainTextReport(report).unicode()
File "/usr/lib/python2.7/site-packages/sos/reporting.py", line 151, in unicode
else six.u(i)), buf))
File "/usr/lib/python2.7/site-packages/sos/reporting.py", line 151, in <lambda>
else six.u(i)), buf))
File "/usr/lib/python2.7/site-packages/six.py", line 647, in u
return unicode(s.replace(r'\\', r'\\\\'), "unicode_escape")
UnicodeDecodeError: 'unicodeescape' codec can't decode byte 0x5c in position 25: \ at end of string
Avoid this problem by re-packing the buffer and adding a single ' '
character to the end of any string ending with '\'.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
| |
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
$ make rpm
rm: remove write-protected regular file './sos/sosreport.pyc'? y
rm: remove write-protected regular file './sos/policies/__pycache__/debian.cpython-35.pyc'? ^CMakefile:74: recipe for target 'clean' failed
make: *** [clean] Interrupt
Gets old very fast..
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
| |
The iteritems() member has been removed in python3. Since the
environment loop in sos_get_command_output() is not expected to
be run with very large inputs use .items() instead, which is
compatible with both python2 and python3.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
| |
For compatibility with python3, when using startswith() or split()
with data read from a file, use bytes rather than a string.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This patch adds to the networking plugin the command:
$ nft list rulesets
Which will provide all currently loaded nftables rulesets. In addition,
it collects static nftables configuration.
Signed-off-by: Robb Manes <rmanes@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Plugin names must match the filename they are found in:
# sosreport -l | grep gnocchi
openstack_gnocchi inactive Gnocchi - Metric as a service
# sosreport -o openstack_gnocchi
sosreport (version 3.3)
a non-existing plugin (openstack_gnocchi) was specified in the command line
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
added command to collect 'ss -lnp' within each namespace
(edited to use -peaonmi for consistence with collection in the
main namespace. -- bmr).
Resolves: #966.
Signed-off-by: Martin Schuppert <mschuppert@redhat.com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adds the following to the networking plugin:
$ ss -peaonmi
The ss tool provides much more detailed socket information than netstat
and is almost guaranteed to be available while netstat not always so.
ss provides some very useful extra TCP socket information such as memory
allocations, the congestion window size, window scaling factor, segments
sent/received, outstanding data, etc, etc. It also provides the socket
address for every known socket type listed.
The net-tools package is not part of the default package set for some
minimal EL installs so it is not always available. ss is part of the
more-modern iproute suite so is more likely to always be available.
Signed-off-by: Patrick Talbert <ptalbert@redhat.com>
|
|
|
|
|
|
| |
Resolves: #914
Signed-off-by: Jacob Wen <jian.w.wen@oralce.com>
|
|
|
|
|
|
|
|
|
| |
_get_npm_output with working_directory=project_path must be called only if
project_path is configured
Resolves: #951
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
| |
Don't try to open corosync.conf if it doesn't exist, and handle
exceptions on open and IO if it does.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
| |
Resolves: #934.
Signed-off-by: Eric Desrochers <eric.desrochers@canonical.com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
| |
This patch enables sosreport to capture files under /sys/fs/pstore/ .
Signed-off-by: Mukesh Ojha <mukesh02 linux vnet ibm com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
| |
This patch enables sosreport to collect "diag_encl -v" command output.
Signed-off-by: Mukesh Ojha <mukesh02 linux vnet ibm com>
Signed-off-by: Ankit Kumar <ankit linux vnet ibm com>
Tested-by: Mukesh Ojha <mukesh02 linux vnet ibm com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
lsvpd family commands captures information, which can be extracted from
/var/lib/lsvpd. So, avoid capturing duplicate information by skipping
the execution of these command.
Signed-off-by: Mukesh Ojha <mukesh02 linux vnet ibm com>
Reviewed-by: Kamalesh Babulal <kamalesh linux vnet ibm com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The change in how commands are parsed was introduced in commit
04d412816de2e4053f43956adf536272a2ddc19c. This caused our old dpkg parsing
to not work. Whatever the reason for the \* in the command it doesn't appear to
be needed now anyway so removing it.
Resolves: #935
Signed-off-by: Bryan Quigley <bryan.quigley@canonical.com>
|
|
|
|
|
|
| |
Resolves: #972.
Signed-off-by: Aaron Conole <aconole@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
dpdk_nic_bind was renamed by the DPDK project to dpdk_devbind, which
means we need to test for either.
Additionally, driverctl is used in rhel systems to provide udev based
persistent overrides, so capturing its output is helpful as well.
Signed-off-by: Aaron Conole <aconole@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Specify packages plugin variable to run the plugin only on systems
with gnocchi installed.
Resolves: #977
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
| |
Catch IOError when opening /etc/collectd.conf and log it as a warning.
Resolves: #968, #976.
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
| |
Set transactions = -1 to skip for loop when dnf history is empty.
Resolves: #969
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
|
| |
"import os.path" missing.
Resolves: #967
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
| |
Pass the file object in which HTML and text reports have been
generated to the archiving class. Since the files received are
instances of TempFileUtil files, they will use fdopen and cannot
return a valid file name.
|
|
|
|
|
|
|
|
| |
Use a StreamHandler for sos.log and ui.log so that data is sent
directly to the log files, and to avoid the FileHandler creating
a bogus '<fdopen>' directory entry.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Open files created by the TempFileUtil class for both reading and
writing: this allows the log files that are managed by this class
to be read back once they have been written.
This will be used to convert logging handlers to StreamHandler.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Allow subclasses of FileCacheArchive to accept either a path to
a file to add, or an instance of an open File object, from whih
data will be read.
This will be used to covert logging FileHandler object to
StreamHandler objects, avoiding creating bogus '<fdopen>' paths
in the file system when the former are used with files that are
created with os.fdopen().
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If auth information got sourced before sosreport run (OS_USERNAME,
OS_PASSWORD, OS_TENANT_NAME) collect information from nova, neutron,
keystone, glance, ironic and heat.
Only undercloud specific heat information is collected in instack
plugin.
Closes: #955.
Signed-off-by: Martin Schuppert <mschuppe@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Package list shall never timeout, otherwise pkgs gets empty, causing
KeyError later on.
Resolves: #942
Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
|
|
|
|
|
|
|
|
|
| |
The current add_copy_spec_limit() implementation does not correctly
support recursive directory addition: to make it more clear that
we are collecting everything from the manila log directory use a
'*' at the end of the path spec.
Signed-off-by: Bryn M. Reeves <bmr@redhat.com>
|