[Impala-ASF-CR] IMPALA-7550: Add documentation to profile counters

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/14776 )

Change subject: IMPALA-7550: Add documentation to profile counters
..


Patch Set 8: Code-Review+1

Thanks Jiawei, LGTM.


--
To view, visit http://gerrit.cloudera.org:8080/14776
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Idc03faddb27754001290bb6d899840e2cbe7ccb7
Gerrit-Change-Number: 14776
Gerrit-PatchSet: 8
Gerrit-Owner: Jiawei Wang 
Gerrit-Reviewer: Andrew Sherman 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: David Rorke 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Reviewer: Jiawei Wang 
Gerrit-Reviewer: Jiawei Wang 
Gerrit-Reviewer: Lars Volker 
Gerrit-Reviewer: Tim Armstrong 
Gerrit-Comment-Date: Thu, 12 Dec 2019 10:55:21 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-7550: Add documentation to profile counters

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/14776 )

Change subject: IMPALA-7550: Add documentation to profile counters
..


Patch Set 5:

(28 comments)

Looks nice. I haven't looked at the significance fields for individual counters 
yet. In general, all but DEBUG descriptions should avoid Impala-internal 
terminology as much as possible. The target audience will never want to open 
Impala's code, so referencing function names, etc. will be frustrating. Some 
are harder to word around than others (e.g. scan range), I skipped those for 
now.

http://gerrit.cloudera.org:8080/#/c/14776/5//COMMIT_MSG
Commit Message:

http://gerrit.cloudera.org:8080/#/c/14776/5//COMMIT_MSG@37
PS5, Line 37: 2. Profile counters are annotated with their stability:
: * Stable counters - generally useful to understand query 
performance,
: should only change rarely and if it does we'll make some effort to
: notify users. E.g. BytesRead.
: * Unstable but useful - useful to understand query performance, 
but
: subject to change, particularly if the implementation changes. 
E.g.
: RowBatchQueuePutWaitTime, MaterializeTupleTimer
: * Debugging counters - generally not useful to users of Impala, 
the main
: use case is low-level debugging. Can be hidden to reduce noise 
for most
: consumers of profiles.
:
: 3. Profile counters are also annotated with their significance to 
users.
: * Critical level counters - always useful on measuring query 
performance and status.
: Counters that everyone are interested.
: * High level counters - generally interesting counters. Most of 
the users will be
: interested and all the developers are very interested.
: * Medium level counters - somehow interesting counters to 
monitor. It will probably be
: interesting under some circumstance. Lot of developers are 
interested.
: * Low level counters - not interesting to users. Should be useful 
for developers
: to debug only.
Please simplify this to what's in the code - I think the descriptions in 
Significance are a good explanation of the different levels.


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc
File be/src/exec/hdfs-scan-node-base.cc:

http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@70
PS5, Line 70: by Disk I/O threads in HDFS read operations. For example, if we 
have 3 reading "
: "threads and each spent 1 sec, this counter will report 3 
sec.")
Instead of including an explanation in individual description (but not in e.g. 
the next one), I'd stick to a common language that's clear enough by itself for 
all 'wall clock timers over multiple threads' descriptions. Maybe 'Aggregate 
wall clock time across all Disk I/O threads...'?


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@79
PS5, Line 79: concept of \"scan range\", including HDFS and Kudu.
Is this included in profiles where it's irrelevant? If no, remove this to avoid 
confusion. If yes, it'd be better to enumerate the full list since 'scan range' 
is not a user-facing term.


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@85
PS5, Line 85: disk queue
Maybe '...remote data source...'? Disk queue is not a term users will be 
familiar with. Or, again, just enumerate.
Is HDFS remote read overall a single disk queue, or individual remote read 
targets? Same for S3.


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@88
PS5, Line 88: Higher values
Than what? - Is there a point of comparison we can provide? For example, 
'values close to the number of disks accessed'?


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@89
PS5, Line 89: thread
scan


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@90
PS5, Line 90: because of
I don't think we should speculate on root causes - there can be many reasons 
for this.


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@121
PS5, Line 121: traditional HDFS scan nodes and the scan "
 : "node total time for the MT_DOP > 1 scan nodes
Isn't this the same as 'HDFS scans'?


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@122
PS5, Line 122: Low values show
Don't scanner threads shut down if there's no work to do? IIUC, slow reads 
result in low ScannerIoWaitTime and also low AverageNumScannerThreads. If 
that's correct, I'd just focus on interpreting high values.


http://gerrit.cloudera.org:8080/#/c/14776/5/be/src/exec/hdfs-scan-node-base.cc@125
PS5, Line 125: Note that if CPU load is high, this "
 : "can 

[Impala-ASF-CR] IMPALA-8729: [DOCS] Describe on-demand metadata feature

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/13802 )

Change subject: IMPALA-8729: [DOCS] Describe on-demand metadata feature
..


Patch Set 1: Code-Review+1

(2 comments)

LGTM overall

http://gerrit.cloudera.org:8080/#/c/13802/1/docs/topics/impala_metadata.xml
File docs/topics/impala_metadata.xml:

http://gerrit.cloudera.org:8080/#/c/13802/1/docs/topics/impala_metadata.xml@61
PS1, Line 61: and cache it locally.
Since you mention lack of eviction in the problem statement, I think we could 
mention here that the impalad's cache gets evicted automatically under memory 
pressure.


http://gerrit.cloudera.org:8080/#/c/13802/1/docs/topics/impala_metadata.xml@65
PS1, Line 65: metadata fetches
This is true between the coordinator and catalog only. The catalogd still has 
to load entire tables to be able to serve any request.



--
To view, visit http://gerrit.cloudera.org:8080/13802
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I64261625c1d9b122c7cca59f9b004dda05810351
Gerrit-Change-Number: 13802
Gerrit-PatchSet: 1
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Bharath Vissapragada 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Comment-Date: Fri, 05 Jul 2019 07:55:12 +
Gerrit-HasComments: Yes

[Impala-ASF-CR] IMPALA-8224: [DOCS] Add the missing contents about Impala web UI pages

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/12934 )

Change subject: IMPALA-8224: [DOCS] Add the missing contents about Impala web 
UI pages
..


Patch Set 2: Code-Review+1

LGTM, sorry for the delay.


--
To view, visit http://gerrit.cloudera.org:8080/12934
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ic30c26fe1d19c1a81704723cd5ebed667982d5b6
Gerrit-Change-Number: 12934
Gerrit-PatchSet: 2
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Andrew Sherman 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Bikramjeet Vig 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Reviewer: Sahil Takiar 
Gerrit-Comment-Date: Tue, 09 Apr 2019 07:40:45 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-7107: [DOCS] Updated the doc for storage formats impala cannot insert into

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/12529 )

Change subject: IMPALA-7107: [DOCS] Updated the doc for storage formats impala 
cannot insert into
..


Patch Set 2: Code-Review+1

LGTM. Thanks!


--
To view, visit http://gerrit.cloudera.org:8080/12529
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: If154e24ade83e2cde2b08878e6f4f88e0048d657
Gerrit-Change-Number: 12529
Gerrit-PatchSet: 2
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Reviewer: Zsombor Fedor (406)
Gerrit-Comment-Date: Thu, 28 Feb 2019 21:36:47 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-6741: Add timestamp of fragment instance's status updates

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/12000 )

Change subject: IMPALA-6741: Add timestamp of fragment instance's status updates
..


Patch Set 2: Code-Review+1

LGTM


--
To view, visit http://gerrit.cloudera.org:8080/12000
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Iae3dcddc292d694d7003d10ed0caccfceed7d8fa
Gerrit-Change-Number: 12000
Gerrit-PatchSet: 2
Gerrit-Owner: Michael Ho 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Reviewer: Lars Volker 
Gerrit-Reviewer: Michael Ho 
Gerrit-Reviewer: Zoram Thanga 
Gerrit-Comment-Date: Thu, 29 Nov 2018 12:52:25 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-5826 IMPALA-7162: [DOCS] Documented the IDLE SESSION TIMEOUT query option

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/11004 )

Change subject: IMPALA-5826 IMPALA-7162: [DOCS] Documented the 
IDLE_SESSION_TIMEOUT query option
..


Patch Set 2: Code-Review+1


--
To view, visit http://gerrit.cloudera.org:8080/11004
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I37182a3c5cf19fdcbb5f247ed71d43f963143510
Gerrit-Change-Number: 11004
Gerrit-PatchSet: 2
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Reviewer: Tim Armstrong 
Gerrit-Reviewer: Zoltan Borok-Nagy 
Gerrit-Comment-Date: Mon, 23 Jul 2018 20:57:34 +
Gerrit-HasComments: No

[Impala-ASF-CR] [DOCS] Clarification on admission control and DDL statements

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10829 )

Change subject: [DOCS] Clarification on admission control and DDL statements
..


Patch Set 3:

Thanks Tim for clarifying the session-thing. I was thinking in impala-shell 
context only.


--
To view, visit http://gerrit.cloudera.org:8080/10829
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I2e3e82bd34e88e7a13de1864aeb97f01023bc715
Gerrit-Change-Number: 10829
Gerrit-PatchSet: 3
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Reviewer: Tim Armstrong 
Gerrit-Comment-Date: Tue, 03 Jul 2018 10:55:44 +
Gerrit-HasComments: No

[Impala-ASF-CR] [DOCS] Clarification on admission control and DDL statements

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10829 )

Change subject: [DOCS] Clarification on admission control and DDL statements
..


Patch Set 1:

I think this is somehow much more confusing than it should be. The facts are:
* queries submitted to a given session are executed serially, DDL or not
* DDLs are not part of admission control
* Naturally, this means that any query, including DDLs, that the client submits 
in a session that has a query queued, will not execute (but this behaviour has 
nothing to do with admission control). The tricky part is that while the 
example query is queued, it is possible to open a new session and drop the 
table the queued query would refer to.

I think we should have a page somewhere (probably here?) to separate these 
behaviours (queued due to serial exec vs queued due to admission control), and 
spell out how this relates to DDLs specifically (the scenario above is a good 
example IMO).


--
To view, visit http://gerrit.cloudera.org:8080/10829
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I2e3e82bd34e88e7a13de1864aeb97f01023bc715
Gerrit-Change-Number: 10829
Gerrit-PatchSet: 1
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Comment-Date: Tue, 26 Jun 2018 21:48:33 +
Gerrit-HasComments: No

[Impala-ASF-CR] [DOCS] Fixed a typo for missing 'not'

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10657 )

Change subject: [DOCS] Fixed a typo for missing 'not'
..


Patch Set 2: Code-Review+1


--
To view, visit http://gerrit.cloudera.org:8080/10657
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I8342bcb47d4a9aa422e234e488dd1dfbdc1694d4
Gerrit-Change-Number: 10657
Gerrit-PatchSet: 2
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Comment-Date: Mon, 18 Jun 2018 16:55:22 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-7100: [DOCS] Consistent memory alloc across executor nodes

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10561 )

Change subject: IMPALA-7100: [DOCS] Consistent memory alloc across executor 
nodes
..


Patch Set 2: Code-Review+1

Feel free to carry the +1 if you've addressed the comments.


--
To view, visit http://gerrit.cloudera.org:8080/10561
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I22926eb6050f9501624d2041f594fe4ef15be73b
Gerrit-Change-Number: 10561
Gerrit-PatchSet: 2
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Comment-Date: Tue, 05 Jun 2018 20:33:53 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-7050: [DOCS] Document the max serialized incremental stat size setting

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10457 )

Change subject: IMPALA-7050: [DOCS] Document the max serialized incremental 
stat size setting
..


Patch Set 3:

(7 comments)

initial pass

http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_perf_stats.xml
File docs/topics/impala_perf_stats.xml:

http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_perf_stats.xml@709
PS3, Line 709: When you need to run COMPUTE INCREMENTAL STATS 
on
 : very large tables, you can use the configuration 
setting
 :   inc_stats_size_limit_bytes to 
reduce the load on
 : the catalog server.
turn into passive?


http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_perf_stats.xml@714
PS3, Line 714: This limit is set as a safety check, to prevent the JVM from 
hitting
 : a maximum array limit of 1 GB or runs out of memory.
the limit is 2GB with suggested JVM versions. Readers would need to understand 
that this is only a part of the entire table's metadata (all of which together 
must be below 2GB when serialized as Thrift). So while we're trying to avoid a 
2GB limit, increasing the 200MB limit can easily lead to crashes (other parts 
of the table's metadata can add up quickly).


http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_perf_stats.xml@724
PS3, Line 724: To change the inc_stats_size_limit_bytes value
This is a startup option for the impala daemons and the catalogd, not a client 
(impala-shell) side config.


http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_perf_stats.xml@733
PS3, Line 733: 1 GB
I'd put 500MB


http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_perf_stats.xml@733
PS3, Line 733: spike in heap usage
as well as crashes.


http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_perf_stats.xml@736
PS3, Line 736: 
 : Setting inc_stats_size_limit_bytes 
to a big value,
 : such as 1 GB or more, can result in a spike in heap 
usage.
 :   
duplicate of above?


http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_shell_options.xml
File docs/topics/impala_shell_options.xml:

http://gerrit.cloudera.org:8080/#/c/10457/3/docs/topics/impala_shell_options.xml@570
PS3, Line 570: --inc_stats_size_limit_bytes
this is an impalad-side option



--
To view, visit http://gerrit.cloudera.org:8080/10457
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ifa80325f0008d42a9cc8178e7c144fc2b49d7d4e
Gerrit-Change-Number: 10457
Gerrit-PatchSet: 3
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins 
Gerrit-Comment-Date: Tue, 05 Jun 2018 20:31:20 +
Gerrit-HasComments: Yes

[Impala-ASF-CR] IMPALA-6714: [DOCS] ORC file format support

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10525 )

Change subject: IMPALA-6714: [DOCS] ORC file format support
..


Patch Set 2: Code-Review+1

LGTM


--
To view, visit http://gerrit.cloudera.org:8080/10525
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ib1ee23ed844653c274babdce5a332dbe5c79b630
Gerrit-Change-Number: 10525
Gerrit-PatchSet: 2
Gerrit-Owner: Quanlong Huang 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Michael Brown 
Gerrit-Reviewer: Quanlong Huang 
Gerrit-Comment-Date: Mon, 04 Jun 2018 14:02:44 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-6714: [DOCS] ORC file format support

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10525 )

Change subject: IMPALA-6714: [DOCS] ORC file format support
..


Patch Set 1:

(2 comments)

http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml
File docs/topics/impala_orc.xml:

http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml@93
PS1, Line 93: If you do not have an existing data file to use, begin by 
creating one in the appropriate format.
> OK. This is the same as other formats' docs. Do you think they should all b
Yea don't think this is very helpful. Thanks for pointing out this is all over, 
created IMPALA-7107.


http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml@152
PS1, Line 152: Enabling Compression for ORC Tables
> I think it's reasonable. There're no details examples in the official site
You're right. This comes a bit closer: 
https://cwiki.apache.org/confluence/display/Hive/LanguageManual+ORC#LanguageManualORC-HiveQLSyntax,
 but still lacks examples. My concern is that this documents Hive behaviour, 
not Impala (e.g Hive might change the preferred way of altering compression). I 
don't feel strongly about this, we can cover it in IMPALA-7107 if need be.



--
To view, visit http://gerrit.cloudera.org:8080/10525
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ib1ee23ed844653c274babdce5a332dbe5c79b630
Gerrit-Change-Number: 10525
Gerrit-PatchSet: 1
Gerrit-Owner: Quanlong Huang 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Michael Brown 
Gerrit-Reviewer: Quanlong Huang 
Gerrit-Comment-Date: Fri, 01 Jun 2018 13:13:28 +
Gerrit-HasComments: Yes

[Impala-ASF-CR] IMPALA-6714: [DOCS] ORC file format support

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/10525 )

Change subject: IMPALA-6714: [DOCS] ORC file format support
..


Patch Set 1:

(6 comments)

Thanks for doing this!

http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_file_formats.xml
File docs/topics/impala_file_formats.xml:

http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_file_formats.xml@115
PS1, Line 115: parquet
orc


http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_file_formats.xml@124
PS1, Line 124: Before that, create the table using Hive.
Remove - before 2.12, Impala won't be able to query anyway, right?


http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml
File docs/topics/impala_orc.xml:

http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml@93
PS1, Line 93: If you do not have an existing data file to use, begin by 
creating one in the appropriate format.
The example below should be enough, remove.


http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml@152
PS1, Line 152: Enabling Compression for ORC Tables
This section deals mostly with Hive - is there a Hive document that could be 
referenced instead including the commands?


http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml@260
PS1, Line 260: Most of the types have the same name in Impala except the BINARY 
type is STRING type in Impala,
 : and the DATE type is not supported in Impala.
Turn into list (or box, similar to what Parquet has)


http://gerrit.cloudera.org:8080/#/c/10525/1/docs/topics/impala_orc.xml@269
PS1, Line 269: For example,
Add examples of what works, and one which doesn't. Include exception text.



--
To view, visit http://gerrit.cloudera.org:8080/10525
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ib1ee23ed844653c274babdce5a332dbe5c79b630
Gerrit-Change-Number: 10525
Gerrit-PatchSet: 1
Gerrit-Owner: Quanlong Huang 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Comment-Date: Tue, 29 May 2018 09:28:05 +
Gerrit-HasComments: Yes

[Impala-ASF-CR] IMPALA-6553: [DOCS] load catalog in background default change

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/9389 )

Change subject: IMPALA-6553: [DOCS] load_catalog_in_background default change
..


Patch Set 4: Code-Review+1

LGTM, after looking around commit comments are often omitted for docs changes.


--
To view, visit http://gerrit.cloudera.org:8080/9389
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I548b2d1532c12f8d3c795a940b7f980482ecf09b
Gerrit-Change-Number: 9389
Gerrit-PatchSet: 4
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins
Gerrit-Reviewer: John Russell 
Gerrit-Reviewer: Tim Armstrong 
Gerrit-Comment-Date: Thu, 22 Feb 2018 21:55:24 +
Gerrit-HasComments: No

[Impala-ASF-CR] IMPALA-6553: [DOCS] load catalog in background default change

[Balazs Jeszenszky (Code Review)]

Balazs Jeszenszky has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/9389 )

Change subject: IMPALA-6553: [DOCS] load_catalog_in_background default change
..


Patch Set 3:

(3 comments)

Add commit comment?

http://gerrit.cloudera.org:8080/#/c/9389/3/docs/shared/impala_common.xml
File docs/shared/impala_common.xml:

http://gerrit.cloudera.org:8080/#/c/9389/3/docs/shared/impala_common.xml@3465
PS3, Line 3465: Both on startup and after a global or a table-specific metadata
  : invalidate, the metadata load for tables that 
actual queries can
  : get backed up because the catalog service could 
spend time
  : loading metadata for another table. During such 
load, a request
  : for prioritized load by a query will be put on 
hold,
This should be simpler.
Consider: 'Background load can interfere with query-specific metadata loading. 
This can happen on startup or after invalidating metadata, with a duration 
depending on the amount of metadata [making it difficult to diag]'


http://gerrit.cloudera.org:8080/#/c/9389/3/docs/shared/impala_common.xml@3470
PS3, Line 3470: random long-running queries that are difficult
  : for you to diagnose.
remove 'for you'


http://gerrit.cloudera.org:8080/#/c/9389/3/docs/shared/impala_common.xml@3475
PS3, Line 3475: increasing
potentially increasing



--
To view, visit http://gerrit.cloudera.org:8080/9389
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I548b2d1532c12f8d3c795a940b7f980482ecf09b
Gerrit-Change-Number: 9389
Gerrit-PatchSet: 3
Gerrit-Owner: Alex Rodoni 
Gerrit-Reviewer: Balazs Jeszenszky 
Gerrit-Reviewer: Impala Public Jenkins
Gerrit-Reviewer: John Russell 
Gerrit-Reviewer: Tim Armstrong 
Gerrit-Comment-Date: Thu, 22 Feb 2018 09:57:34 +
Gerrit-HasComments: Yes