summaryrefslogtreecommitdiffstats
path: root/__root__
Commit message (Collapse)AuthorAgeFilesLines
* Add infrastructure for "report bugs" referenceJan Pokorný2016-02-012-0/+2
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc: move rgmanager-pacemaker.* to dedicated directoryJan Pokorný2016-02-016-1/+1
| | | | | | Also rename env_vars -> env-vars (for uniformity). Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* filters/cib2pcscmd[resources]: drop op options "spam"Jan Pokorný2016-02-011-4/+2
| | | | | | Also reflect the changes in the tests. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten: inject also some resource meta-attributes backJan Pokorný2016-02-011-0/+7
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* Haziness: 'monitor' and 'status' are synonymousJan Pokorný2016-02-011-1/+1
| | | | | | | Former is preferred, per: <http://www.linux-ha.org/doc/dev-guides/_actions.html> Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.03.groups: FAILURE-ISOLATION propertyJan Pokorný2016-02-012-1/+68
| | | | | | In very rough state for now. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.03.groups: ENABLED+RECOVERY from resourcesJan Pokorný2016-02-012-137/+207
| | | | | | | | ... as they were actually meant like that from the beginning (it's true that Pacemaker can mimic some of these also with primitive resources, but we are looking at the conversion perspective now). Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.02.resources: prefer kind to score with orderJan Pokorný2016-02-011-2/+7
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.02.resource: internal COOCCURRENCE ref.Jan Pokorný2016-02-011-0/+1
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.02.resources: typos (occurrence, positive)Jan Pokorný2016-02-011-32/+32
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.00.intro: fix discrepanciesJan Pokorný2016-02-011-3/+3
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.02.resources: MANAGED -> ENABLEDJan Pokorný2016-02-011-12/+20
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.02.resources: EXCLUSIVE alternative in PJan Pokorný2016-02-011-0/+4
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.02.resources: pcs supports utilization nowJan Pokorný2016-02-011-1/+2
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* doc/rgmanager-pacemaker.02.resources: adjust relocate recoveryJan Pokorný2016-02-011-3/+3
| | | | | | ... per what's actually implemented. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: be verbose about exit statusJan Pokorný2016-02-011-1/+1
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: fix nonshort short-circuiting + Py2.6 -m pkg.__main__Jan Pokorný2016-02-011-4/+9
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: add nosetests support (also as a fallback)Jan Pokorný2016-02-011-9/+37
| | | | | | (fix tabs/spaces issue along the way) Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-sdist: be more careful about tarball with testsJan Pokorný2016-02-011-3/+7
| | | | | | | | Previously, the fact that --transform applies also to the symlink content was missed and resulted in an archive that wasn't directly usable. Also drop maintenance-specific "vim local rc" file. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-sdist/clufter.spec: fix tests archive arangementJan Pokorný2015-12-181-4/+3
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* filters~ccs2pcs: propagate "monitor" op for primitivesJan Pokorný2015-12-171-1/+6
| | | | | | | | | | | | | | For this monitor operation, either use the interval as designated in the original rgmanager agent (sticking with the most basic=lowest depth only), or use the default of 60 seconds (the same is automatically injected as a default by pcs). Resolves: <https://bugzilla.redhat.com/1272570>. Also adjust the (unit) tests respectively. Reported-by: Miroslav Lisik <mlisik@redhat.com> Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten: inject also resource-specific action stanzasJan Pokorný2015-12-171-1/+32
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-sdist: make it create additional "tests" tarballJan Pokorný2015-12-171-5/+12
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* README: reflect the current (fixed) ccs2pcscmd outputJan Pokorný2015-12-171-3/+4
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* README: fix the ccompiled typoJan Pokorný2015-12-171-1/+1
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* filters/cmd_wrap: cmd_args_cutter: special-case "stonith create"Jan Pokorný2015-12-171-2/+1
| | | | | | Adapt that in the test as well. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: run also run_check_cmd from run-check unless -n passedJan Pokorný2015-12-171-1/+1
| | | | | | This should prevent run_check_cmd getting stale. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* filters/cmd_wrap: revisit group/line wrap-filling algoJan Pokorný2015-12-171-8/+15
| | | | | | | | | | | | | Previous one was just splitting into lines within a single group (also, this was limited to at most one split! that in turn made it a pretty broken text wrapping). Now it is two-ways allowing also short-enough groups to be put on the same line, and as many line splits is made as necessary for a proper text wrapping. Also update the test counterparts (function run_check_cmd of run-check script succeeds again after quite a long time!). Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: fix a bug in logic and case statementsJan Pokorný2015-12-171-3/+6
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* filters/coro2pcscmd[corosync]: fix unresolved cluster nameJan Pokorný2015-12-171-1/+1
| | | | | | | | | | It is in /corosync/totem/@cluster_name for corosync.conf XML representation, not at the location expected in cluster.conf (/cluster/@name). Also fix the (still failing) ccs2pcscmd test run in run-check. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* README: add Coveralls badgesJan Pokorný2015-12-171-0/+6
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: coverage.py under Travis CI doesn't recognize unit2 cmdJan Pokorný2015-12-171-1/+1
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: account also for cases when there is no "[["Jan Pokorný2015-12-171-1/+1
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: account for cases when there is no coverage2Jan Pokorný2015-12-171-1/+2
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-{check,tests}: fix "."/passed params dash incompatibilityJan Pokorný2015-12-172-2/+3
| | | | | | (this one was affecting Travis CI runs) Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-tests: add support for running under coverage.pyJan Pokorný2015-12-172-2/+11
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* KNOWN-ISSUES: finally start tracking themJan Pokorný2015-12-171-1/+6
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-check: --dist goes prior alias cmd to influence itJan Pokorný2015-10-081-4/+4
| | | | | | | | | | | | | | | | This hasn't been considered before (hence the bug being fixed), and it seems best if --dist in "global options" (prior to cmd spec) is the only to influence command alias resolution, whereas --dist as "command option" can influence some specifics for the command itself. This way, natural causality on the command line is preserved. Alternatively, we could look-ahead if there is dist defined as command option and considered it in the alias resolution. But rather than that, it would more bearable to remove --dist/--sys from command options altogether but that would prevent injecting them when command-by-symlink is exercised, so do not do this either. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-check: add a (sanity) check for ccs2pcscmd outputJan Pokorný2015-09-091-1/+77
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* run-check: we can rely on dist-implied alias resolutionJan Pokorný2015-09-091-2/+2
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten: fix possible memleakJan Pokorný2015-09-091-18/+25
| | | | | | | | | | | | | | | | | | | | | | | | | ...found by static analysis Account for simplistic logic of the call site (our goal is not "read as many bytes as you can, stop on first allocation error, but be fine with what was read so far, if any" -- that is not intended when you need to read whole parseable XML). There are more considerations: - realloc(p, 0) may or may not free p, one shouldn't free it manually to prevent double-free (do I understand "man 3p realloc" right?) - length (size_t) can overflow, leading realloc to actually decrease the allocated memory, perhaps even to zero, i.e., possible free as mentioned above) - realloc(NULL, size) == malloc(size) - check in "if (x) free(x);" is redundant - if read_pipe returns OK, we can now safely assume: size > 0 Also following principles are followed: - there should be just one cleanup on error - notable code branches and otherwise silent assumptions commented upon Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten/resrules.c: fix errno.h include in wrong filEJan Pokorný2015-09-092-1/+1
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten: inject also primary attribute nameJan Pokorný2015-09-023-0/+22
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten: remove metadata for long deprecated tomcat-5 RAJan Pokorný2015-09-021-104/+0
| | | | | | See <https://bugzilla.redhat.com/show_bug.cgi?id=593721>. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten: accidentally omitted SAP{Database,Instance} metadataJan Pokorný2015-09-022-0/+171
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* ccs-flatten/Makefile: agents do not always have trailing .shJan Pokorný2015-09-021-2/+4
| | | | | | Cf. SAP{Database,Instance}. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* clufter.spec: prepare grounds for possible pagure.io jumpJan Pokorný2015-09-021-1/+6
| | | | | | Open source infrastructure FTW! Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* filters/cibprelude2cibcompact[resources]: fix rule+stickinessJan Pokorný2015-08-111-1/+2
| | | | | | | | | Apparently, rule is supposed to accompany the actual nvpair that will be applied in case such condition holds (otherwise it's a no-op). Adjust a check script respectively. Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* setup: include "void"/minimal format instances (files)Jan Pokorný2015-07-141-0/+1
| | | | Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
* README: expand ccs_flatten dependencies/requirementsJan Pokorný2015-07-021-2/+6
| | | | | | (merge build- and run-time deps) Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
generated by cgit (git 2.34.1) at 2026-04-22 00:14:43 +0000