Hi Kunal,

Thanks for the response. I have created JIRA 
DRILL-6183<https://issues.apache.org/jira/browse/DRILL-6183> as you mentioned.


Regards,


Arjun

________________________________
From: Kunal Khatua <kunalkha...@gmail.com>
Sent: Thursday, February 22, 2018 12:09 PM
To: user@drill.apache.org
Subject: Re: Default value for parameter 'planner.width.max_per_node'

Looks like DRILL-5547: Linking config options with system option manager ( 
https://github.com/apache/drill/commit/a51c98b8bf210bbe9d3f4018361d937252d1226d 
) introduced a change in the computation, which is based on the number of cores.
[https://camo.githubusercontent.com/ef395746e2290c22e7998b2ded24bc569babaf4f/68747470733a2f2f302e67726176617461722e636f6d2f6176617461722f39653833376539326635303030643735653831376533356230646664373131613f643d68747470732533412532462532466173736574732d63646e2e6769746875622e636f6d253246696d6167657325324667726176617461727325324667726176617461722d757365722d3432302e706e6726723d6726733d323030]<https://github.com/apache/drill/commit/a51c98b8bf210bbe9d3f4018361d937252d1226d>

DRILL-5547: Linking config options with system option manager ยท 
apache/drill@a51c98b<https://github.com/apache/drill/commit/a51c98b8bf210bbe9d3f4018361d937252d1226d>
github.com
closes #868



https://github.com/apache/drill/blame/6ad9838339999c3e069b6925d0c3ce3077afd8c3/exec/java-exec/src/main/java/org/apache/drill/exec/server/options/TypeValidators.java#L252
[https://avatars3.githubusercontent.com/u/47359?s=400&v=4]<https://github.com/apache/drill/blame/6ad9838339999c3e069b6925d0c3ce3077afd8c3/exec/java-exec/src/main/java/org/apache/drill/exec/server/options/TypeValidators.java#L252>

apache/drill<https://github.com/apache/drill/blame/6ad9838339999c3e069b6925d0c3ce3077afd8c3/exec/java-exec/src/main/java/org/apache/drill/exec/server/options/TypeValidators.java#L252>
github.com
drill - Mirror of Apache Drill




Looks like a non-zero value is not being computed. Setting the value as a 
system variable should persist and correct this. Can you file a JIRA for this? 
The fix should be minor.

On 2/21/2018 8:54:53 PM, Arjun kr <arjun...@outlook.com> wrote:
Hi Team,


Did default value for configuration parameter 'planner.width.max_per_node' 
change between Drill 1.10 and 1.11 versions? Also, does default value of 70% of 
total cores still hold true?



-- Drill 1.11


0: jdbc:drill:drillbit=localhost> select * from sys.options where name like 
'%planner.width%';
+------------------------------+-------+-------------------+--------------+----------+----------+-------------+-----------+------------+
| name | kind | accessibleScopes | optionScope | status | num_val | string_val 
| bool_val | float_val |
+------------------------------+-------+-------------------+--------------+----------+----------+-------------+-----------+------------+
| planner.width.max_per_node | LONG | ALL | BOOT | DEFAULT | 0 | null | null | 
null |
| planner.width.max_per_query | LONG | ALL | BOOT | DEFAULT | 1000 | null | 
null | null |
+------------------------------+-------+-------------------+--------------+----------+----------+-------------+-----------+------------+
2 rows selected (0.913 seconds)
0: jdbc:drill:drillbit=localhost>

-Drill 1.10

0: jdbc:drill:schema=dfs> select * from sys.options where name like 
'%planner.width%';
+------------------------------+-------+---------+----------+----------+-------------+-----------+------------+
| name | kind | type | status | num_val | string_val | bool_val | float_val |
+------------------------------+-------+---------+----------+----------+-------------+-----------+------------+
| planner.width.max_per_node | LONG | SYSTEM | DEFAULT | 3 | null | null | null 
|
| planner.width.max_per_query | LONG | SYSTEM | DEFAULT | 1000 | null | null | 
null |
+------------------------------+-------+---------+----------+----------+-------------+-----------+------------+
2 rows selected (0.359 seconds)
0: jdbc:drill:schema=dfs>

Thanks,

Arjun

Reply via email to