[
https://issues.apache.org/jira/browse/PHOENIX-1982?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14550879#comment-14550879
]
James Taylor edited comment on PHOENIX-1982 at 5/19/15 5:56 PM:
----------------------------------------------------------------
Nice, [~rajeshbabu]! Here's some minor feedback:
- I'd recommend changing the below to "As of now, users need to manually
add/delete the UDF jars into/from HDFS. In a future release, we will support
these operations through new Phoenix SQL statements."
{code}
+<b><em>As of now user need to add/delete the udf jars into/from HDFS. In
upcoming versions
+we can going to support these operations through queries.</em></b>
{code}
- I'd recommend changing this slightly to "You can follow these simple steps to
write your UDF (for more detail, see
[this](http://phoenix-hbase.blogspot.in/2013/04/how-to-add-your-own-built-in-function.html)
blog post):":
{code}
+You can following simple steps to write custom or domain-specific function.
These steps required here have been
+picked from blog post written by James
http://phoenix-hbase.blogspot.in/2013/04/how-to-add-your-own-built-in-function.html.
+You can refer it if any doubts.
{code}
- Slight addition here, "Additionally, to enable an ORDER BY to be optimized
out or a GROUP BY to be done in-place,"
{code}
+* Additionally, to maintain an optimization for GROUP BY to be done in-place,
+we need to override the following method:
{code}
- Maybe add a Limitations section that mentions existing JIRAs like 1) adding
new SQL statements for add/remove jars, 2) functional indexes need to manually
be rebuilt if the implementation changes, 3) Once loaded, a jar will not be
unloaded, so you'll need to put modified implementations into a different jar
to prevent having to bounce your cluster. 4) ...
- Don't forget to add a link in site.xml under the Features menu to the new UDF
page.
was (Author: jamestaylor):
Nice, [~rajeshbabu]! Here's some minor feedback:
- I'd recommend changing the below to "As of now, users need to manually
add/delete the UDF jars into/from HDFS. In a future release, we will support
these operations through new Phoenix SQL statements."
{code}
+<b><em>As of now user need to add/delete the udf jars into/from HDFS. In
upcoming versions
+we can going to support these operations through queries.</em></b>
{code]
- I'd recommend changing this slightly to "You can follow these simple steps to
write your UDF (for more detail, see
[this](http://phoenix-hbase.blogspot.in/2013/04/how-to-add-your-own-built-in-function.html)
blog post):":
{code}
+You can following simple steps to write custom or domain-specific function.
These steps required here have been
+picked from blog post written by James
http://phoenix-hbase.blogspot.in/2013/04/how-to-add-your-own-built-in-function.html.
+You can refer it if any doubts.
{code}
- Slight addition here, "Additionally, to enable an ORDER BY to be optimized
out or a GROUP BY to be done in-place,"
{code}
+* Additionally, to maintain an optimization for GROUP BY to be done in-place,
+we need to override the following method:
{code}
- Maybe add a Limitations section that mentions existing JIRAs like 1) adding
new SQL statements for add/remove jars, 2) functional indexes need to manually
be rebuilt if the implementation changes, 3) Once loaded, a jar will not be
unloaded, so you'll need to put modified implementations into a different jar
to prevent having to bounce your cluster. 4) ...
- Don't forget to add a link in site.xml under the Features menu to the new UDF
page.
> Documentation for UDF support
> -----------------------------
>
> Key: PHOENIX-1982
> URL: https://issues.apache.org/jira/browse/PHOENIX-1982
> Project: Phoenix
> Issue Type: Sub-task
> Reporter: Rajeshbabu Chintaguntla
> Assignee: Rajeshbabu Chintaguntla
> Fix For: 4.4.0
>
> Attachments: PHOENIX-1982.patch, create_function.csv
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
