| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
(merge build- and run-time deps)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
| |
Unfortunately changing the order of ccs2pcs_needle kwargs also changed
the order in which the arguments (when parsed in a terse/option-less
notation) are expected --> fix that in run-check script as well.
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
| |
Hopefully there is no major roadblock making this change unreasonable at
this point in time (checked commonly used RHEL, Fedora, Debian/Ubuntu
-- the latter also for seamless integration with Travis CI).
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
| |
See, e.g.,
https://git.fedorahosted.org/cgit/cluster.git/commit/?h=STABLE32&id=5bddd1e21098a183560f64a569dd15fa8fdca1a8
https://bugzilla.redhat.com/show_bug.cgi?id=707118#c9
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
| |
... as this is the most suitable moment (we will not be parsing
resource rules/metadata later again).
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Hence hard-code the distro to the testing command to make sure it
reproduces what the hardwired referential results where obtained with.
Check, e.g., this commit to learn about what I mean:
filters/ccsflat2cibprelude[clusternode]: conditional node id.
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
(marked suspicious by ShellCheck tool)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
| |
... that also makes this wrapped iteration more-or-less compatible with
list_remove (see the in-line comment, but discretion needed, untested!)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
list_do: do not check *list twice
list_done: do not check curr, as we are quite sure it will never reach
NULL value implicitly (using intended list manipulators incl.
the wrapped iteration itself, not explicit redefinition that
is broken anyway), otherwise we would have already observed
either trivial infloops or segfaults caused by reliance on
such an assumption (curr != NULL)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
Also drop unnecessary assert.h include from xmlconf.c
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
(code marked suspicious by static analysis)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
(found by static analysis)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
(finding of static analysis)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
Discovered thanks to Travis CI tests.
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
| |
We need this sort of determinism (not needed generally as agents are
supposed to be different, otherwise it would confuse cluster management,
etc.).
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, running the CLI without proper installation on the machine
with some /usr/share/cluster/*.metadata files present (typically RHEL 6
with resource-agents installed) meant that "onvenient fallback for
local/test deployment" hadn't been activated, leading to various errors
in the conversion due to most of the *.metadata files missing anyway
(most notably for "script").
Now we take an advantage of own/embedded metadata files ending with
".sh.metadata", therefore we are not confused in this way (at least
not so easily, but good enough for the time being).
Reported-by: Chris Feist <cfeist@redhat.com>
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
|
| |
- add "#ifdef RA_METADATA_EXT" where also advisable
- closedir before possibly returning, not the other way around
- reflect the fact basename doesn't return NULL and work
with its return value directly
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
| |
- new: bind-mount, db2
- updated: nfsserver, postgres-8, vm
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
| |
...hence monkey-patching monkey-patched (by setuptools against distutils'
original implementation)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
Also include this file into the specfile as extra source if appropriate.
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
...when figuring out the "own identity"
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
| |
(harmless since basename function is idempotent)
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
| |
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|
|
|
|
|
|
|
|
|
| |
Also delete run-setup as it has about the same meaning now as running
"python setup.py" fixed part of the command directly.
See also: https://github.com/pypa/pip/pull/2337
Signed-off-by: Jan Pokorný <jpokorny@redhat.com>
|