What causes replication to fail with ERROR 1037 (HY001): Out of memory; please try the operation again?
The error indicates that you are trying to replicate a database from one cluster to another cluster where the database is too big to fit in the secondary cluster.
memsql> replicate database database_name from user:'XXXXX@<IP of prod server>3306;
ERROR 1037 (HY001): Out of memory; please try the operation again (needed ****** bytes)
Check the following to fix the issue:
- Compare the cluster sizing of both primary and secondary clusters. Next, we need to ensure that the Storage, Memory, and CPU configurations are equal on both clusters.
- Check for System misconfiguration at Linux kernel/OS Level; click here to learn more.
- If Numa is available on Primary Cluster, Configure the secondary cluster for the Numa. Click here to learn more.
- Check and fix the memory misconfiguration (maximum_memory, maximum_table_memory) on the nodes; click here to learn more.
Also, Cluster issues can be verified using the sdb-report toolbox command below; click here to learn more about the tool.
Click here to learn about Replicating Data Across Clusters.