Hi,
You are right. The memory of this node is small, but the data volume is
large. There is data exchange between memory and disk during SQL query,
which we think is normal.
But according to iostat's statistics,
r/s 2852 *4k = 11.4M~rMB/s
That is, each time the disk reads data, it only reads about 4KB
(pageSize) of data? Is this very inefficient? Not one block of data at a
time?
Or do I need to optimize some parameter?
在 2020/5/13 上午3:51, akorensh 写道:
Hi,
Attach the logs and I'll take a look. Include the details of your sql
request: # of rows/size of objects/
indexs/sql queries/etc..
It might be using the disk and swapping memory back and forth.
https://apacheignite.readme.io/docs/durable-memory-tuning#pages-writes-throttling
Use this guide to tune memory:
https://apacheignite.readme.io/docs/durable-memory-tuning
more optimizations here:
https://apacheignite.readme.io/docs/performance-tips
https://apacheignite.readme.io/docs/preparing-for-production
Thanks, Alex
--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/
