Sahil Takiar has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/14314 )

Change subject: IMPALA-8826: [DOCS] Add docs for PLAN_ROOT_SINK and result 
spooling
......................................................................


Patch Set 1:

(12 comments)

http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_fetch_rows_timeout_ms.xml
File docs/topics/impala_fetch_rows_timeout_ms.xml:

http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_fetch_rows_timeout_ms.xml@21
PS1, Line 21: IMPALA-3200
is this the right JIRA?


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_fetch_rows_timeout_ms.xml@46
PS1, Line 46: </p>
should mention IMPALA-8962 as well. the timeout also applies to fetch request 
issued against 'RUNNING' queries. a query that is 'RUNNING' has no rows 
available, so any fetch request will wait until the query transitions to the 
'FINISHED' state and for it to fetch all requested rows. a query in the 
'FINISHED' state means that the rows are available to be fetched (the query may 
actually still be executing though).


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_fetch_rows_timeout_ms.xml@51
PS1, Line 51: 10
default is 10 seconds, so 10000 ms, should the units be somewhere?


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_max_result_spooling_mem.xml
File docs/topics/impala_max_result_spooling_mem.xml:

http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_max_result_spooling_mem.xml@39
PS1, Line 39: unpinned
I don't think our docs really explain the difference between unpinned vs pinned 
memory, so we should probably avoid mentioning pinned and unpinned memory 
altogether.

for this section, we can just say that all memory will most likely be spilled 
to disk.


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_max_spilled_result_spooling_mem.xml
File docs/topics/impala_max_spilled_result_spooling_mem.xml:

http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_max_spilled_result_spooling_mem.xml@47
PS1, Line 47: to allow unpinning all pinned
            :       memory when the amount of spooled results exceeds
            :         <codeph>MAX_RESULT_SPOOLING_MEM</codeph>
can just remove this part


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml
File docs/topics/impala_query_results_spooling.xml:

http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml@32
PS1, Line 32: cause
causing


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml@32
PS1, Line 32: wait for
            :         extended
wait for an extended


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml@40
PS1, Line 40: this spooled locations
the spooled results.


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml@54
PS1, Line 54: unpinned
remove unpin reference here and below


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml@95
PS1, Line 95: of the number of rows returned by the query
of the amount of memory needed to spool all the rows returned by the query


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml@98
PS1, Line 98: and
remove "and". the read and write buffers are 2 MB in size each, which is why 
the default is 4 MB


http://gerrit.cloudera.org:8080/#/c/14314/1/docs/topics/impala_query_results_spooling.xml@117
PS1, Line 117: asynchronously
remove, it isn't really asynchronously



--
To view, visit http://gerrit.cloudera.org:8080/14314
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I78bfceb225d25078c54c1ed8f88ca250ef42dafe
Gerrit-Change-Number: 14314
Gerrit-PatchSet: 1
Gerrit-Owner: Alex Rodoni <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Sahil Takiar <[email protected]>
Gerrit-Comment-Date: Tue, 01 Oct 2019 00:12:46 +0000
Gerrit-HasComments: Yes

Reply via email to