FrankChen021 opened a new issue #9899:
URL: https://github.com/apache/druid/issues/9899


   this issue is related to #9755 . At first I thought it was because there 
were too many rows in the `druid_segments` table. So we killed unused segments 
through API to reduce the records in that table. After that, the count of rows 
dropped from about 360,000 to about 50,000.
   
   But our DBA still reported that there were slow queries there, which pointed 
to statement `SELECT payload FROM druid_segments WHERE used=true`. So I ran 
this SQL from local mysql client, and it reported that it took more than 90 
seconds to get the result.
   
   I ran another SQL `SELECT id FROM druid_segments WHERE used=true` from 
local, which reported that it took more than 12 seconds to complete.
   
   Since the `payload` field in database is a `LONGBLOB`, I guess the reason 
why the query is so slow is because the return set is too large in bytes, which 
takes a long time to transmit data from MySQL server to client in network.
   
   ### Affected Version
   
   the version we currently use is 0.16.0. since this issue is related to 
database, Ii think the latest version is still affected
   
   ### Description
   
   
![image](https://user-images.githubusercontent.com/6525742/82414101-1f176980-9aa9-11ea-90ed-d3d5afdc432a.png)
   
   
   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
[email protected]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to