Hi There
I think I probably missed something, or mis-understood about Tablespace.
I'm new to Cluster but used the "traditionnal" Mysql server for some years.
My issue is there are some tables I would like to transfer to disk storage.
When I create a tablespace however it seems to be using the server's RAM rather than storing the file to disk ? Is that possible.
For example I created 1 log file and 1 tablespace as follow:
CREATE LOGFILE GROUP lg_1
ADD UNDOFILE 'undo_1.log'
INITIAL_SIZE 536870912
UNDO_BUFFER_SIZE 16777216
ENGINE NDBCLUSTER;
ALTER LOGFILE GROUP lg_1
ADD UNDOFILE 'undo_2.log'
INITIAL_SIZE 314572800
ENGINE NDBCLUSTER;
CREATE TABLESPACE ts_1
ADD DATAFILE 'data_1.dat'
USE LOGFILE GROUP lg_1
INITIAL_SIZE 1073741824
EXTENT_SIZE 1048576
ENGINE NDBCLUSTER;
That seems to increase the RAM usage of the server by about 1.5GB.
This is a test installation on this server with 1 SQL node, 1 Management node and 2 Data nodes on the same machine. For production of course we'll spread across different physical machines.
I am using Cluster v 5.1.51-ndb-7.1.9a-cluster-gpl-log x64
The OS is Win 2008 Server x64 with 16GB of RAM and dual quad core Xeon
Is that normal behavior or did I miss one or many parameters ?
Many thanks in advance for your help.
Vincent
I think I probably missed something, or mis-understood about Tablespace.
I'm new to Cluster but used the "traditionnal" Mysql server for some years.
My issue is there are some tables I would like to transfer to disk storage.
When I create a tablespace however it seems to be using the server's RAM rather than storing the file to disk ? Is that possible.
For example I created 1 log file and 1 tablespace as follow:
CREATE LOGFILE GROUP lg_1
ADD UNDOFILE 'undo_1.log'
INITIAL_SIZE 536870912
UNDO_BUFFER_SIZE 16777216
ENGINE NDBCLUSTER;
ALTER LOGFILE GROUP lg_1
ADD UNDOFILE 'undo_2.log'
INITIAL_SIZE 314572800
ENGINE NDBCLUSTER;
CREATE TABLESPACE ts_1
ADD DATAFILE 'data_1.dat'
USE LOGFILE GROUP lg_1
INITIAL_SIZE 1073741824
EXTENT_SIZE 1048576
ENGINE NDBCLUSTER;
That seems to increase the RAM usage of the server by about 1.5GB.
This is a test installation on this server with 1 SQL node, 1 Management node and 2 Data nodes on the same machine. For production of course we'll spread across different physical machines.
I am using Cluster v 5.1.51-ndb-7.1.9a-cluster-gpl-log x64
The OS is Win 2008 Server x64 with 16GB of RAM and dual quad core Xeon
Is that normal behavior or did I miss one or many parameters ?
Many thanks in advance for your help.
Vincent