[
https://issues.apache.org/jira/browse/PHOENIX-5728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17039696#comment-17039696
]
Viraj Jasani commented on PHOENIX-5728:
---------------------------------------
Thanks [~gjacoby]
> wouldn't it be simpler in many cases to just have an object-based explain
>plan with individual fields that can be asserted against?
I agree with your opinion. For the comparison of individual fields, do you
recommend json with detailed info for each field and tests can convert the json
to a POJO that can be easy to assert? If I am not mistaken, I can start with
working on explain plan and as we discover more fields to
serialize/deserialize, we can add more field in the POJO.
What might be difficult is extracting individual fields of our interest from
string e.g. "CLIENT 1-CHUNK 0 ROWS 20 BYTES PARALLEL 1-WAY FULL SCAN OVER %s\n
SERVER FILTER BY FIRST KEY ONLY", from this one, we need to extract Client,
Rows, Bytes, Scan, Filter etc.
I believe you are suggesting this improvement specifically because as of now we
have just List<String> in ExplainPlan. Correct?
> Gold files for Phoenix test BaseStatsCollectorIT
> ------------------------------------------------
>
> Key: PHOENIX-5728
> URL: https://issues.apache.org/jira/browse/PHOENIX-5728
> Project: Phoenix
> Issue Type: Sub-task
> Affects Versions: 5.1.0, 4.15.1
> Reporter: Viraj Jasani
> Assignee: Viraj Jasani
> Priority: Major
> Fix For: 5.1.0, 4.15.1
>
> Attachments: PHOENIX-5728.master.000.patch,
> PHOENIX-5728.master.000.patch
>
>
> Provide golden files for QueryPlan result comparison for all abstract test
> class BaseStatsCollectorIT that includes tests of all stats collections
> including:
> * NamespaceDisabledStatsCollectorIT
> * NamespaceEnabledStatsCollectorIT
> * NonTxStatsCollectorIT
> * TxStatsCollectorIT
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
