Change in vdsm[master]: virt: introduce support for periodic operations
fromani at redhat.com
fromani at redhat.com
Tue Mar 31 06:25:21 UTC 2015
Francesco Romani has posted comments on this change.
Change subject: virt: introduce support for periodic operations
......................................................................
Patch Set 29:
(1 comment)
https://gerrit.ovirt.org/#/c/37593/29/tests/periodicTests.py
File tests/periodicTests.py:
Line 144: def _work():
Line 145: invokations[0] += 1
Line 146: invokations[1] = monotonic_time()
Line 147: if invokations[0] == BLOCK_AT:
Line 148: # must be > (PERIOD * TIMES) ~= forever
> may I ask why you opted against an explicit done.wait() in this thread?
Sure, unfortunately the answer is in the realm of personal preferences, but here it is:
IMO using time.sleep(), the code can be grokked in a single pass. It is straightforward to see what we do and why we do it that way.
Using done.wait() made me stop for a single tiny moment to think. Indeed was a tiny moment, but was a hiccup anyway: why is the same function using set() and wait() on the same event?!
For this reason I kept sleep(), which also mimics a little better the real world - an operation can be still blocked when we shut down things.
As I said first, this is 99% personal taste, so I have no real objections to switch to done()
Line 149: time.sleep(10 * PERIOD * TIMES)
Line 150: executions[0] += 1
Line 151: executions[1] = monotonic_time()
Line 152: if invokations[0] == TIMES:
--
To view, visit https://gerrit.ovirt.org/37593
To unsubscribe, visit https://gerrit.ovirt.org/settings
Gerrit-MessageType: comment
Gerrit-Change-Id: I834c067b1d3dba16a1f8c83e555d2cc04cbf32ed
Gerrit-PatchSet: 29
Gerrit-Project: vdsm
Gerrit-Branch: master
Gerrit-Owner: Francesco Romani <fromani at redhat.com>
Gerrit-Reviewer: Dan Kenigsberg <danken at redhat.com>
Gerrit-Reviewer: Francesco Romani <fromani at redhat.com>
Gerrit-Reviewer: automation at ovirt.org
Gerrit-HasComments: Yes
More information about the vdsm-patches
mailing list