[
https://issues.apache.org/jira/browse/IOTDB-736?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17125571#comment-17125571
]
Julian Feinauer commented on IOTDB-736:
---------------------------------------
[~hxd]I think the easiest thing then would be to define a separate option for a
query-logger and just write it to a separate logfile. If I add the Metrics
logging in there (See IOTDB-23) then I can also do that I think.
> Query performance tracing
> -------------------------
>
> Key: IOTDB-736
> URL: https://issues.apache.org/jira/browse/IOTDB-736
> Project: Apache IoTDB
> Issue Type: New Feature
> Components: Core/Engine
> Reporter: Xiangdong Huang
> Priority: Major
>
> Hi,
> Today I run a `select count(s1) from root.some_time_series`, while there is
> no un-sequence data. The data type is text and there is 1.5TB data totally
> and the time series has about 15GB data, then it spends 17s.
>
> is that possible to trace the query process for debugging the performance?
> Like what Cassandra provides, `tracing on`, and then how many sstables or
> tombstone are read and the time cost is clear.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
