I have 4 column store tables with 11 million records in one database. Each database occupied around 380mb. Like this I have created 82 orgs.
We are using the memsql from AWS market palce 4:1 cluster. When I check my memsql-studio, I can see as each leaf nodes 23gb ram full out off 25 gb & 710 gb of hard disk is full out of 1tb hard disk.
1)Why column store tables are occupied ram and hard disk. Is it expected behavior?
select database_name,
table_name,
format(sum(memory_use),0) memory_in_bytes,
format(sum(memory_use)/(1024),0) memory_in_kb,
format(sum(memory_use)/(10241024),0) memory_in_mb,
format(sum(memory_use)/(10241024*1024),0) memory_in_gb
from information_schema.table_statistics
group by 1, 2;
2)When I ran the above query data is showing as no memory occupied. How to check through a query how much memory and disk occupied the database tables?
3)I am unable to load the data further. I am receiving the error below.
SQL Error [1712] [HY000]: Leaf Error(10.100.0.18:3306): The operating system failed to allocate memory
4)We are going to store multiple organizations’ data in same database. In order to store multitenant data we need to estimate how much ram and disk required. Is there any mechanism you advise to us.