Edit your config.init file and allocate more OperationRecords (one is
needed per row in transaction).
This is a hard limit at this point in time, but will be modified in
Matteo Brancaleoni wrote:
>I'm doing some tests with the cluster, now succesfully
>running onto 2 servers, one with the MGM,DB1 and API,
>the other with DB2.
>I was running a test script that DELETE * FROM table,
>then inserts 10k records, then do a count, and go
>on from start.
>after a while I got the temp error 233 from ndbcluster
>"Out of operation records in transaction coordinator"
>restarting the api (mysql server) or the cluster itself
>What that could mean?