Hi guys,
Just pushing that pmie change to my private tree for now - if you
need it Lukas, its just a git-pull away. Mark, could you also take
a look at the pmiostat change there & pull if all good?
thanks!
Changes committed to git://git.pcp.io/nathans/pcp.git master
qa/312 | 2 +-
qa/518 | 8 ++++----
src/pmdas/linux_proc/help_text.h | 2 +-
src/pmie/src/pmie.c | 4 ++--
src/pmiostat/pmiostat.py | 15 ++++++++-------
5 files changed, 16 insertions(+), 15 deletions(-)
commit d21bb6c017696233d736d0c518eeeb742eb4d36f
Author: Nathan Scott <nathans@xxxxxxxxxx>
Date: Tue Apr 21 16:24:40 2015 +1000
pmie: remove unnecessary 2 second delay at live startup
Noticed that pmie was taking a while to start up and dump out the
result from initial sample for qa312. ltrace pointed toward some
1 second calls to nanosleep, and following that led to unexpected
time deltas for scheduled tasks vs the initial notion of "now".
Turned out those delays were due to some places where we calcualte
'sched_time = now + 1sec' (or similar) - seems to be unnecessary,
from QA testing so far.
commit 8048bf8cb660583cf5e9f95682b214930f0ababa
Author: Nathan Scott <nathans@xxxxxxxxxx>
Date: Tue Apr 21 16:21:57 2015 +1000
pmiostat: usage message cleanup
Noticed pmiostat --help was printing "None" right at the
end, which seemed odd. Turned out to be a not-quite-right
API call in the usage exception handler.
While there, found the 'x' option not being presented via
the getopt API call (harmless in practice), a double-colon
on "General Options::", and some whitespace/alignment not
quite in the right spot - so, all trivial cleanups here.
commit e0e0a7a879fff29414867e65c9a18096b3811cf2
Author: Nathan Scott <nathans@xxxxxxxxxx>
Date: Tue Apr 21 16:21:15 2015 +1000
qa: fix short-help-text length test on proc.psinfo.environ
|