On Wed, 2012-07-11 at 13:49 -0300, Lucas Meneghel Rodrigues wrote: > On Wed, Jul 11, 2012 at 1:46 PM, Lucas Meneghel Rodrigues > <look...@gmail.com> wrote: > > On Tue, Jul 10, 2012 at 10:16 AM, <sathn...@linux.vnet.ibm.com> wrote: > >> From: Satheesh Rajendran <sathn...@linux.vnet.ibm.com> > >> > >> This test implements the following github issue > >> https://github.com/autotest/autotest/issues/404 > > > > Hi Satheesh, this test looks generally good to me, but there's one > > gotcha, see below: > > > > diff --git a/client/virt/guest-os.cfg.sample > > b/client/virt/guest-os.cfg.sample > >> index 5c1eb2b..4e19c42 100644 > >> --- a/client/virt/guest-os.cfg.sample > >> +++ b/client/virt/guest-os.cfg.sample > >> @@ -34,6 +34,10 @@ variants: > >> time_command = date +'TIME: %a %m/%d/%Y %H:%M:%S.%N' > >> time_filter_re = "(?:TIME: \w\w\w )(.{19})(?:\.\d\d)" > >> time_format = "%m/%d/%Y %H:%M:%S" > >> + time_manage: > >> + time_command = date +'TIME: %a %m/%d/%Y %H:%M:%S.%N' > >> + time_filter_re = "(?:TIME: \w\w\w )(.{19})(?:\.\d\d)" > >> + time_format = "%m/%d/%Y %H:%M:%S" > >> file_transfer: > >> tmp_dir = /tmp/ > >> clean_cmd = rm -f > >> @@ -1415,6 +1419,11 @@ variants: > >> time_command = "echo TIME: %date% %time%" > >> time_filter_re = "(?<=TIME: \w\w\w ).{19}(?=\.\d\d)" > >> time_format = "%m/%d/%Y %H:%M:%S" > >> + time_manage: > >> + alive_test_cmd = systeminfo > >> + time_command = "echo TIME: %date% %time%" > >> + time_filter_re = "(?<=TIME: \w\w\w ).{19}(?=\.\d\d)" > >> + time_format = "%m/%d/%Y %H:%M:%S" > >> guest_s4: > >> check_s4_support_cmd = powercfg /hibernate on > >> test_s4_cmd = start ping -t localhost > >> diff --git a/client/virt/subtests.cfg.sample > >> b/client/virt/subtests.cfg.sample > >> index 099f28e..875e464 100644 > >> --- a/client/virt/subtests.cfg.sample > >> +++ b/client/virt/subtests.cfg.sample > >> @@ -1026,6 +1026,18 @@ variants: > >> drift_threshold = 10 > >> drift_threshold_single = 3 > >> > >> + - time_manage: > >> + type = time_manage > >> + kill_vm = yes > >> + extra_params +=" -rtc base=utc,driftfix=slew -snapshot" > >> + host_load_command = stress --vm 1 --vm-bytes 4G --vm-keep > >> + host_load_kill_command = killall stress > > > > ^ Here, this is fine *if* stress is installed in the system, so here > > it's necessary to at least put a comment saying that this program > > should be available in $PATH. (moreover, I'm not sure whether this is > > installable in other types of host, such as Debian, Ubuntu, OpenSUSE. > > In any case, this is easy enough to fix, I'll test this a bit more and > > then apply it with the additional comment.
> Another thing I forgot to comment: Here you seem to imply that stress > is only memory pressure, rather than also do IO and CPU stress as > well. Why so? Yes, this test needs to be demonstrated in a more stressed environment, I should have used cpu and io stress too. Once the patch is available in next, I would change them. Thanks, -Satheesh. _______________________________________________ Autotest mailing list Autotest@test.kernel.org http://test.kernel.org/cgi-bin/mailman/listinfo/autotest
