[ 
https://issues.apache.org/jira/browse/CASSANDRA-11483?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15227220#comment-15227220
 ] 

Yuki Morishita commented on CASSANDRA-11483:
--------------------------------------------

Thanks for the update.

Couple of comments:

* I think you can just move CommandLine / Options and static block to 
initialize those inside {{main}}.
* I like to have just command line option without arg. All options right now 
are boolean, and doing like {{-s true}} to activate them is verbose.
* I really like new presentation but some users may be using it for just 
numbers. If you can create API for display and put new representation and old 
one there, that would be great. In the future, we can plugin more presentation 
style (like beloved json).
* Some fields like "repaired at" or "deletion time" have default or special 
value to indicate certain state. For example "Repaired at" == 0 means SSTable 
is never repaired. I think displaying date string to such field can mislead 
user. WDYT?
* Can you clean up import statements and unused variables? Also, some fields 
are using {{System.out}} and others are using instance variable {{out}}.

> Enhance sstablemetadata
> -----------------------
>
>                 Key: CASSANDRA-11483
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-11483
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Observability
>            Reporter: Chris Lohfink
>            Assignee: Chris Lohfink
>            Priority: Minor
>             Fix For: 3.x
>
>         Attachments: CASSANDRA-11483.txt, CASSANDRA-11483v2.txt, 
> CASSANDRA-11483v3.txt, Screen Shot 2016-04-03 at 11.40.32 PM.png
>
>
> sstablemetadata provides quite a bit of useful information but theres a few 
> hiccups I would like to see addressed:
> * Does not use client mode
> * Units are not provided (or anything for that matter). There is data in 
> micros, millis, seconds as durations and timestamps from epoch. But there is 
> no way to tell what one is without a non-trival code dive
> * in general pretty frustrating to parse



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to