Repository navigation
addBatch method cause Java heap memory char array grows too fast #1873
Description
Activity
Hi @wangbing505
- What is the ClickHouse jdbc version do you use?
- Is it a requirement to use JDBC?
Hi @chernser
- ClickHouse jdbc version is 0.2.6，Can a higher version solve this problem？
- yes，or Is there any other way to write data to clickhouse ?
- Yes, please, update to the latest 0.7.0 version. I do not see such code in it and everything should work fine.
- Yes, you may use java client without JDBC - https://clickhouse.com/docs/en/integrations/java/client-v2
Please let me know how the update goes. Thanks!
@chernser
Thank you for your answer，I will try using them and synchronize the resultsReacted by Sergey Chernov@wangbing505 as the issue solved?
@chernser hello, I have been delayed by other things and will try it next week
@chernser Thank you, I tried it, Char [] has decreased，But there are other memory uses as well，maybe it's not a problem with the client anymore
Reacted by Sergey ChernovNeed to verify how jdbc-v2 handles it.
We might need to stream big SQL requests like we do with text formats.
Will prioritize this work.
The scenario involves a large number of calls to the addBatch method, Then use jmap to view heap memory，has lots of char[],although they are all unreachable objects that can be GC dropped.But it will lead to frequent GC and high CPU usage. How to solve it