Michael Brown has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/9770 )

Change subject: IMPALA-5721,IMPALA-6717,IMPALA-6738: improve stress test binary 
search
......................................................................


Patch Set 7:

(7 comments)

http://gerrit.cloudera.org:8080/#/c/9770/1/tests/stress/concurrent_select.py
File tests/stress/concurrent_select.py:

http://gerrit.cloudera.org:8080/#/c/9770/1/tests/stress/concurrent_select.py@174
PS1, Line 174:           LOG.warn(
> docstring
Done


http://gerrit.cloudera.org:8080/#/c/9770/1/tests/stress/concurrent_select.py@176
PS1, Line 176:               "The stress test algorithm needs control of this 
option.".format(
             :                   query_option=query_option, opt=args_common_quer
> Ideally we should also prefix query.logical_id. QueryReport doesn't own tha
Done


http://gerrit.cloudera.org:8080/#/c/9770/1/tests/stress/concurrent_select.py@708
PS1, Line 708: ry")
> Stale var of some older designs. Doesn't seem needed.
Nevermind, actually used now.


http://gerrit.cloudera.org:8080/#/c/9770/1/tests/stress/concurrent_select.py@1475
PS1, Line 1475:     if old_required_mem_mb_with_spilling:
              :       mem_limit = old_required_mem_mb_with_spilling
              :       old_required_mem_mb_with_spilling = None
              :     else:
> Getting a little untenable. Maybe instead just save what we want instead of
Eh, withdrawn. The save list is still longer.


http://gerrit.cloudera.org:8080/#/c/9770/7/tests/stress/concurrent_select.py
File tests/stress/concurrent_select.py:

http://gerrit.cloudera.org:8080/#/c/9770/7/tests/stress/concurrent_select.py@113
PS7, Line 113: class StressArgCurator(object):
> Nit: Is "curate" really the right word here? I think of the act of curating
Done


http://gerrit.cloudera.org:8080/#/c/9770/7/tests/stress/concurrent_select.py@141
PS7, Line 141:       if hasattr(self, curate_method_name):
> So, for what it's worth, rather that having these magic _curate_ methods, t
Nice idea. I tend to think __getattr__ is just as magical, and I tend to avoid 
using it, but I think the more magical one is __getattribute__. But you raise a 
good point and using __getattr__ means I don't need to do the dispatch. Since 
you said you'll be reviewing in batches, I'm going to think on this one a while 
before addressing it one way or another.


http://gerrit.cloudera.org:8080/#/c/9770/7/tests/stress/concurrent_select.py@145
PS7, Line 145:         curated_args[k] = v
> Nit: Might want to explicitly mention in either the class or function docst
Done. I tried to convey this L116-122, but it wasn't enough. I've tried again. 
Please let me know if it can be made clearer.



--
To view, visit http://gerrit.cloudera.org:8080/9770
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I33d036ec93df3016cd4703205078dbdba0168acb
Gerrit-Change-Number: 9770
Gerrit-PatchSet: 7
Gerrit-Owner: Michael Brown <mi...@cloudera.com>
Gerrit-Reviewer: David Knupp <dkn...@cloudera.com>
Gerrit-Reviewer: Michael Brown <mi...@cloudera.com>
Gerrit-Reviewer: Nithya Janarthanan <njanartha...@cloudera.com>
Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com>
Gerrit-Comment-Date: Tue, 27 Mar 2018 21:28:29 +0000
Gerrit-HasComments: Yes

Reply via email to