Re: While changing engine to ndbcluster there is an error The Table is Full

[Şükrü ÖZDEMİR]

Thanks for your reply.

The default datamemory and index memory was 80M and 18M. When was them low the count was 48 and now it is 3G and more but still the count is same.

Do you think, could I wrote the datamemory and index memory wrong?

I have changed them and started nodes with

ndb_mgmd -f /var/lib/mysql-cluster/config.ini --initial --configdir=/var/lib/mysql-cluster/


Mikael Ronström Wrote:
-------------------------------------------------------
> Hi,
>
> This table is a temporary table created while
> performing
> an ALTER TABLE.
>
> Table full here simply means that you run out of
> DataMemory.
>
> So the first 47 tables filled the DataMemory.
>
> Your config has only 3GB of DataMemory, you
> probably
> need a lot more than that. You can probably
> decrease the
> IndexMemory.
>
> So problem you are facing is simply not enough
> memory
> to alter all tables. Remember that MySQL Cluster
> stores
> columns in memory by default (There is an option
> to store
> non-index columns on disk, so adding STORAGE DISK
> to the
> ALTER TABLE can put more of the data on disk.
>
> Rgrds Mikael
>