Change in vdsm[master]: tests: Fix picklecopy random failures

[Code Review]

From Dan Kenigsberg :

Dan Kenigsberg has posted comments on this change.

Change subject: tests: Fix picklecopy random failures
......................................................................

Patch Set 2:

(1 comment)

https://gerrit.ovirt.org/#/c/69738/2//COMMIT_MSG
Commit Message:

Line 3: AuthorDate: 2017-01-05 20:31:06 +0200
Line 4: Commit:     Nir Soffer 
Line 5: CommitDate: 2017-01-05 20:42:18 +0200
Line 6: 
Line 7: tests: Fix picklecopy random failures
The existing test was intentional. In Francesco's original patch, he stated that we must have picklecopy much much faster. Apparently it's not only faster. http://gerrit.ovirt.org/34989

We can change our mind.

But it should be clear that this patch does not fix picklecopy, it simply relaxes the requirement from it. Hence I think title should change.
Line 8: 
Line 9: We have a benchmark for picklecopy, failing randomly on travis, when
Line 10: picklecopy was less than twice faster compared with deepcopy. I don't
Line 11: think it is useful to fail the test randomly because of this reason, and