aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* [foreman] get conditionally list of smart-proxies featuresPavel Moravec2021-03-041-5/+22
| | | | | | | | | | | | | Two changes applied: - drop printing IP address of a smart proxy - add plugopt foreman.proxyfeatures that will trigger collecting features of all smart proxies Closes: #2424 Resolves: #2433 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [man] Multiple fixes in man pagesJose Castillo2021-03-043-8/+20
| | | | | | | | | | | This patch fixes references to sosreport, to the preferred 'sos report'. Also adds "SEE ALSO" consistently for all man pages, and fixes a MAINTAINER line. Resolves: #2432 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [ata] Add "smartcl -l scterc dev-name- commandAkshay Gaikwad2021-03-041-1/+2
| | | | | | | | | | "-l scterc" prints values and description of the SCT Error Recovery Control settings. Resolves: #2426 Signed-off-by: Akshay Gaikwad <akgaikwad001@gmail.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [block] Add block device timeout specAkshay Gaikwad2021-03-041-1/+3
| | | | | | | | | | | The "/sys/block/sd*/device/timeout" and "/sys/block/hd*/device/timeout" holds block device timeouts value in seconds. Resolves: #2425 Signed-off-by: Akshay Gaikwad <akgaikwad001@gmail.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [sos] Bump version to 4.14.1Jake Hunsaker2021-02-253-4/+7
| | | | | | This commit represents the release of sos-4.1 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [candlepin] call psql with --no-password to avoid waiting on inputEvgeni Golov2021-02-251-1/+2
| | | | | | | | | | | | | | | | | | The candlepin plugin tries to read the PostgreSQL password from /etc/candelpin/candlepin.conf, but will fallback to an empty string if one isn't found. This in turn causes psql to hang indefinitely (or at least until we hit the plugin timeout). Add `--no-password` to the psql command to not prompt for a password if one isn't provided via another source. Related: #2421 Resolves: #2422 Signed-off-by: Evgeni Golov <evgeni@golov.de> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [foreman] call psql with --no-password to avoid waiting on inputEvgeni Golov2021-02-251-1/+1
| | | | | | | | | | | | | | | | The foreman plugin tries to read the PostgreSQL password from /etc/foreman/database.yml, but will fallback to an empty string if one isn't found. This in turn causes psql to hang indefinitely (or at least until we hit the plugin timeout). Add `--no-password` to the psql command to not prompt for a password if one isn't provided via another source. Closes: #2421 Signed-off-by: Evgeni Golov <evgeni@golov.de> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [rear] Update the rear plugin to collect needed files.Ryan Blakley2021-02-241-2/+4
| | | | | | | | | | The plugin didn't collect anything under /var/lib/rear, it also didn't collect anything under /etc/rear/mappings. Resolves: #2420 Signed-off-by: Ryan Blakley <rblakley@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [ovn_central] Fix ovn_central sbctl command executionEdward Hope-Morley2021-02-241-10/+30
| | | | | | | | | | | sbctl commands can only be run on a leader node, so gate the commands from running on non-leader nodes with a SoSPredicate Closes: #2418 Resolves: #2419 Signed-off-by: Edward Hope-Morley <edward.hope-morley@canonical.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [ubuntu] Prefer ua over ubuntu-advantages cmdEric Desrochers2021-02-241-1/+7
| | | | | | | | | | | | | | | | | 'ubuntu-advantage' cli has been shorten to 'ua' in version 19 and then symlink at package level to make them both available for users as a smooth transition. While most stable Ubuntu releases are at version 19 and onwards, others aren't yet. The idea is to prefer 'ua' but fall back to the deprecated 'ubuntu-advantage' cli as needed. Resolves: #2417 Signed-off-by: Eric Desrochers <eric.desrochers@canonical.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [cleaner|hostname] Match uppercase domains for kerberosJake Hunsaker2021-02-221-1/+5
| | | | | | | | | | | | | | | Previously, kerberos realms would not be obfuscated due to the convention that those are written in all uppercase. While the parser would match these, the hostname parser would not obfuscate them due to the use of intermediary dicts that were case sensitive. Now, match this convention and if it is found, obfuscate as a whole domain rather than a potential FQDN with a hostname. Closes: #2254 Resolves: #2416 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [hostname_map] Fix duplicating entries of domainsJake Hunsaker2021-02-221-3/+14
| | | | | | | | | | | | | | | | | | | | | | The previous commit bf622ae was to address an issue where the system's hostname would not be obfuscated if the host plugin was not run for an sos execution (provided that it was previously run at least once to generate the initial obfuscation). While this commit solved that particular problem, it injected the domains into the "intermediary" dicts used by the mapping in a way that meant after a certain number of further subsequent runs without the host plugin, we'd end up duplicating our TLDs. E.G. 'example.com' being obfuscated as 'obfuscatedomain0.com' would eventually produce 'obfuscateddomain0.com.com'. Correct this injection to ensure consistent obfuscation regardless of the number of future executions without the host plugin. Related: #2406 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* Clarify default timeout for all pluginsJose Castillo2021-02-222-2/+4
| | | | | | | | | | | | | | | The default timeout for all plugins is 300, but the global timeout is set as -1 in the option list. This is because this timeout is unset by default. This patch attempts to clarify the -1 exposed to the user when using commands like 'sos report -l', via the option description as well as a note clarifying it in the man page. Closes: #2003 Resolves: #2415 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [apache] skip collecting pulp* logsPavel Moravec2021-02-221-2/+5
| | | | | | | | | Since these are collected in the pulp plugin directly. Resolves: #2413 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [gluster] Add glusterd public keys and status filesJose Castillo2021-02-221-1/+7
| | | | | | | | | | | | | | | | | | | | | This patch helps capture some missing files in the gluster plugin, i.e.: Files inside /var/lib/glusterd/glusterfind, like *.status files, that store the required timestamps, *.pem.pub files, that store ssh public keys. We also need to omit the glusterfind_*_secret.pem, which contains the openssh private key. Files inside /var/lib/glusterd/.keys, that contains *.pem.pub, the ssh public key. Closes: RHBZ#1925035, RHBZ#1925419 Resolves: #2411 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [clean] Allow specifying keywords in a text fileJake Hunsaker2021-02-225-2/+26
| | | | | | | | | | | | | | | This commit adds the ability for users to provide a text file with a list of newline-delimited keywords that should be obfuscated, rather than requiring all keywords be specified either by the `--keywords` option or configuration file settings. Files may be provided via the new `--keyword-file` option which is available to `clean`, `collect`, and `report`. Closes: #2401 Resolves: #2408 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [report] add enabling packages per plugin to the overviewPavel Moravec2021-02-171-1/+5
| | | | | | | | Add list of packages enabling a plugin into plugins_overview.py . Closes: #2404 Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
* [clean] Fix hostnames not being obfuscated when host plugin wasn't runJake Hunsaker2021-02-173-8/+59
| | | | | | | | | | | | | | | | | | | | | It was discovered that if a report was generated without the host plugin being run, then the hostname parser would not obfuscate the localhost hostname or domain at all. This was due to the fact that while the default_mapping values were being read into the parser's map's dataset correctly, they weren't being loaded into the 'intermediary' dicts that the parser uses to separate short names from domain names. Fix this by reading the contents of the map's dataset dict (which is populated by the map file) into those intermediary dicts that the parser uses. Additionally, safeguard against overloading domain and host obfuscation numbers when loading this way by setting the respective counts to the next-highest number after the map loading is complete. Resolves: #2406 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [ds] Add healthcheck tool report for Red Hat Directory ServerMark Reynolds2021-02-171-1/+3
| | | | | | | | | | | | For Directory Server (389-ds-base), add the output from the DS's healthcheck tool. See https://bugzilla.redhat.com/show_bug.cgi?id=1688759 Resolves: #2403 Signed-off-by: Mark Reynolds <mreynolds@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [md] Add "mdadm -E" command outputAkshay Gaikwad2021-02-171-0/+2
| | | | | | | | | | | The "mdadm -E" print contents of the metadata stored in devices. -E applies to devices which are components of an array, where -D applies to whole array. Resolves: #2391 Signed-off-by: Akshay Gaikwad <akgaikwad001@gmail.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [ovn-{central,host}] Extend informationEdward Hope-Morley2021-02-112-0/+14
| | | | | | | | | | | | | Extend the ovn-central and ovn-host plugins to include logs from their respective applications. For ovn-central extend to also include Load_Balancer, ACL and Logical_Switch_Port nbdb tables. Closes: #2409 Resolves: #2410 Signed-off-by: Edward Hope-Morley <edward.hope-morley@canonical.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [scsi] Add support to collect SCSI debugging logsMamatha Inamdar2021-02-111-4/+12
| | | | | | | | | | This patch updates scsi plugin to collect additional logs for SCSI devices Resolves: #2379 Signed-off-by: Mamatha Inamdar <mamatha4@linux.vnet.ibm.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [powerpc] Add support to collect HNV infomationMamatha Inamdar2021-02-112-3/+7
| | | | | | | | | | This patch is to update powerpc plugin to collect Hyper-V Network Virtualization information. Resolves: #2379 Signed-off-by: Mamatha Inamdar <mamatha4@linux.vnet.ibm.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [report] add csv output format to plugins_overview.pyPavel Moravec2021-02-111-21/+63
| | | | | | | | | | | - add CSV output format to the script - add distros and profiles info - fix some formatting issues Resolves: #2405 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [policy] switch to sha256 sumBryan Quigley2021-02-096-37/+24
| | | | | | | | | | | | Move cleaner to ignore both sha256/md5s Move collector to use sha256 as well. Move to get hash from stdout Resolves: #2351 Closes: #1204 Signed-off-by: Bryan Quigley <code@bryanquigley.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [SCLPlugin] Fix SCLPlugin enablement triggersJake Hunsaker2021-02-091-10/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It was found that if a system has `scl` available to it, and has a non-scl package installed for a plugin that subclasses `SCLPlugin`, but the scl packages are not installed, then the plugin will not be enabled even for the non-scl collections. This was due to a too-restrictive check if the plugin subclasses `SCLPlugins` that prevented the normal plugin triggers from being checked. The future of Software Collections is uncertain to the sos project at this time, but our current use of it is only 3 plugins; foreman, postgresql, and redis. As such, rather than splitting out `SCLPlugin` further into a distinct separate subclass of a plugin and then having to deal with how to determine which plugin class should be instantiated if both the scl and non-scl plugin would otherwise be enabled, make the `SCLPlugin` enablement conditions not prevent normal plugin enablement, and add gates to the `_scl` methods provided by `SCLPlugin` to abort any collection attempt if the specified scl is not detected. If, at a later date, the project sees further (new) use of the `SCLPlugin` then we can/will review this approach at that time. Closes: #2407 Resolves: #2412 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [cleaner] Add 'cleaner' alias functionalityJose Castillo2021-02-093-4/+7
| | | | | | | | | | | The option 'cleaner' was not implemented as an alias. This patch enables it for both 'sos report' and 'sos collect'. Resolves: #2405 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [cleaner] Automatically create directory for sos-cleaner default_mappingJose Castillo2021-02-091-0/+5
| | | | | | | | | | | | | | | | | | When /etc/sos/cleaner directory is missing, cleaner fails to store new default_mapping file due to: [cleaner] Could not update mapping config file: [Errno 2] No such file or directory: '/etc/sos/cleaner/default_mapping' This patch adds the code to check the existence of the directory and create it if necessary. Resolves: RHBZ#1923937 Closes: #2393 Resolves: #2397 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [build] Add the new presets module to setup.pyJake Hunsaker2021-02-031-1/+2
| | | | | | | | | | | | | | When presets were split out from policies, they were not added to `setup.py` in the list of packages. As such they would not be included in any builds that relied upon using `setup.py` to generate a source tarball. Add those packages to the list so that they may be included. Closes: #2390 Resolves: #2402 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [ovirt_hosted_engine] Add vm.conf fileAkshay Gaikwad2021-02-031-0/+1
| | | | | | | | | | The "/var/run/ovirt-hosted-engine-ha/vm.conf" is used as hosted-engine configuration. Resolves: #2399 Signed-off-by: Akshay Gaikwad <akgaikwad001@gmail.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [sos] Align plugin options with destination namesJake Hunsaker2021-02-039-20/+22
| | | | | | | | | | | | | This commit alters several option long-form names or destination names to align those values in a sensible way. This serves to not only remove some abiguity in option naming in code, but also to make it so that the "effective options" line logged in every sos execution can be direction copy-pasted as a working command. Closes: #2288 Resolves: #2398 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [yum] Mask the proxy passwords in /etc/yum.repos.d/Jose Castillo2021-02-031-0/+5
| | | | | | | | | | | | This patch makes sure that all passwords specified in the opcion proxy_password in any file inside /etc/yum.repos.d/ ends up masked in the sosreport. Closes: #2394 Resolves: #2396 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [report] add a script to print plugins' overviewPavel Moravec2021-02-031-0/+73
| | | | | | | | | | | Let's have a script that generates an overview of each plugin's activity. I.e. for each plugin, print list of items the plugin collects in add_copy_spec or add_cmd_output and other similar methods. Resolves: #2392 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [suse] Properly instantiate a package managerJake Hunsaker2021-01-291-0/+4
| | | | | | | | | | | | | | | The previous change to package managers removed SuSE's manual settings without actually having it use the new `RpmPackageManager` abstraction. This in turn made almost any sos execution fail on SuSE. Fix that, by instantiating the package manager. Credit to fluxcap1 on GitHub for identifying this flaw and the resolution. Related: #2389 Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [rpm] Capture the output of 'rpm --showrc'Jose Castillo2021-01-291-0/+2
| | | | | | | | | | | | This request adds the output of --showrc to sos so we can verify the default values of options set via rpmrc and macros config files. Resolves: RHBZ#1921496 Resolves: #2388 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [insights] Adding .registered file to Insights client config collectionPaul Wayper2021-01-291-0/+1
| | | | | | | | | | It's useful to collect the `.registered` file as well to see if the Insights client is actually registered (and if so, when). Closes: #2372 Resolves: #2373 Signed-off-by: Paul Wayper <paulway@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [kubernetes] Add cdk.master.auth-webhook to journal collectionFelipe Reyes2021-01-291-0/+12
| | | | | | | | | | | In Ubuntu CDK when the deployed with Keystone authentication the service cdk.master.auth-webhook is deployed to handle that integration this change includes this unit to collect its journal. Resolves: #2387 Signed-off-by: Felipe Reyes <felipe.reyes@canonical.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [kubernetes] Fix CDK related file pathsFelipe Reyes2021-01-291-3/+3
| | | | | | | | | | | The `files` property is used to determined if it's the kubernetes master node is where the plugin is runnig since it's the only node capable of running kubectl. In Ubuntu CDK environments the file that's only available in the master is /root/cdk/cdk_addons_kubectl_config while /root/cdk/kubeproxyconfig is also available in the workers. Signed-off-by: Felipe Reyes <felipe.reyes@canonical.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [systemd] Prefer resolvectl over systemd-resolveMichael Biebl2021-01-291-4/+10
| | | | | | | | | | | The latter is a deprecated compat symlink. See https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=979264 Resolves: #2385 Signed-off-by: Michael Biebl <biebl@debian.org> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [bcache] Add a new plugin for bcachePonnuvel Palaniyappan2021-01-291-0/+56
| | | | | | | | | | | bcache is used as a caching device (typically an SSD) for HDDs; bcache stats are useful to identify performance problems. Closes: #2378 Resolves: #2384 Signed-off-by: Ponnuvel Palaniyappan <ponnuvel.palaniyappan@canonical.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [kdump] Gather the file kexec-dmesg.logJose Castillo2021-01-291-1/+2
| | | | | | | | | | | | | Moved the file name from kexec-kdump.log to the right one, kexec-dmesg.log and added it to the list of files to gather via add_copy_spec as per #1546. Resolves: RHBZ#1817042 Resolves: #2386 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Pavel Moravec <pmoravec@redhat.com>
* [ovirt_node] Updated the ovirt_node plugin to collect additional dataLev Veyde2021-01-271-0/+15
| | | | | | | | | | | The plugin now collects information from certificates that were generated by oVirt on the host(s)/node(s). Related: RHBZ#1845877 Resolves: #2364 Signed-off-by: Lev Veyde <lveyde@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [ovirt] Updated the ovirt plugin to collect additional dataLev Veyde2021-01-271-1/+29
| | | | | | | | | | The plugin now collects information from certificates that were generated by oVirt. Related: RHBZ#1845877 Signed-off-by: Lev Veyde <lveyde@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [networking] collect tc filters per each devicePavel Moravec2021-01-251-1/+2
| | | | | | | | | Collect "tc -s filter show dev <DEV>" for each device. Resolves: #2383 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [networking] collect devlink infoPavel Moravec2021-01-251-0/+8
| | | | | | | | | Collect both devlink info and params, as well as details per each device. Related to: #2383 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [openvswitch] include ipsec debugging informationAaron Conole2021-01-251-2/+10
| | | | | | | | | | Enhancement to include details for the ovs ipsec monitor daemon. Resolves: #2382 Signed-off-by: Aaron Conole <aconole@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [s390] Collect dasd.conf fileJose Castillo2021-01-251-0/+1
| | | | | | | | | | | | | | | | From RHBZ: The /etc/dasd.conf is the common configuration file for the dasd portion of s390, similar to how /etc/multipath.conf is to multipath. It has settings like failover timings, readonly settings, diag settings, etc. It's very useful to see what the customer has set in this file when doing an initial analysis of s390 and more specifically dasd (s390 storage) issues. It's also useful to have when suggesting dasd tuning changes to the customer. Related: RHBZ#1919277 Resolves: #2381 Signed-off-by: Jose Castillo <jcastillo@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [powerpc] Collect opal-prd log filesVasant Hegde2021-01-251-1/+2
| | | | | | | | | Recent distros redirecting opal-prd logs to /var/log/opal-prd.log file. Resolves: #2377 Signed-off-by: Vasant Hegde <hegdevasant@linux.vnet.ibm.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [networking] Collect 'ethtool -e <device>' conditionally onlyPavel Moravec2021-01-201-11/+11
| | | | | | | | | | | EEPROM dump collection might hang on specific types of devices, or negatively impact the system otherwise. As a safe option, sos report should collect the command when explicitly asked via a plugopt only. Resolves: #2376 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
* [report] clean up references to ticket-numberPavel Moravec2021-01-194-11/+1
| | | | | | | | | | | | | Cleaning up references of --ticket-number, as it was fully replaced by --case-id. The credit goes to @mamatha4 . Resolves: #2375 Relates to: #2374 Signed-off-by: Pavel Moravec <pmoravec@redhat.com> Signed-off-by: Jake Hunsaker <jhunsake@redhat.com>
generated by cgit (git 2.34.1) at 2024-11-16 22:27:55 +0000