How to fix ERROR 1297: Got temporary error from NDB in MySQL

The error message format is: ERROR 1297 (HY000): Got temporary error XXX '...' from NDBCLUSTER. Note the XXX error code as it determines the root cause.

Common codes:
- 410: REDO log overload
- 1217: Out of local operation records
- 233: Out of transaction coordinator records
- 4010: Node failure during transaction
- 1218: Send buffer overload

For other codes, consult MySQL NDB Cluster documentation.

Since this is a temporary error, implement retry logic in your application. Use exponential backoff to avoid overwhelming the cluster:

# Python example
import time
max_retries = 5
retry_delay = 1

for attempt in range(max_retries):
    try:
        # Execute your SQL query
        cursor.execute(sql_statement)
        break
    except mysql.connector.Error as err:
        if err.errno == 1297:
            if attempt < max_retries - 1:
                time.sleep(retry_delay)
                retry_delay *= 2  # Exponential backoff
            else:
                raise
        else:
            raise

For error 1297, a single retry often succeeds when the temporary condition resolves.

If the error message shows "Out of operation records", increase operation record limits in config.ini on the NDB management node:

[ndbd default]
MaxNoOfConcurrentOperations=100000
MaxNoOfLocalOperations=100000
MaxNoOfConcurrentTransactions=16384

Default values are often too low for high-concurrency workloads. After editing config.ini, restart all NDB data nodes:

# Stop NDB cluster
ndb_mgm -e "STOP"

# Restart data nodes
ndbd --reload-config

# Restart SQL nodes
systemctl restart mysql

If error 410 (REDO log overload) appears, adjust checkpointing and log file settings in config.ini:

[ndbd default]
TimeBetweenLocalCheckpoints=20
NoOfFragmentLogFiles=16
FragmentLogFileSize=256M

Increase NoOfFragmentLogFiles or reduce TimeBetweenLocalCheckpoints to complete checkpoints more frequently. These changes take effect after cluster restart.

If error 1218 (send buffer overload) occurs, increase NDB network buffer allocation in config.ini:

[ndbd default]
SendBufferMemory=4M
TotalSendBufferMemory=256M

Larger send buffers prevent network congestion under high query rates. The TotalSendBufferMemory should be 2-4 MB × number of data nodes.

Insufficient DataMemory causes multiple error 1297 variants. In config.ini, set DataMemory to at least 250MB or higher based on your dataset:

[ndbd default]
DataMemory=1024M
IndexMemory=512M

Check current usage in NDB management console:

ndb_mgm
> ALL REPORT MEMORY

If DataMemory is close to 100% utilized, increase it and restart the cluster.

After adjusting configuration, monitor the cluster for error recurrence:

# Check cluster status
ndb_mgm -e "SHOW"

# View data node logs
tail -f /var/log/mysql/ndb_XX_cluster.log

# Monitor for ERROR 1297
sudo journalctl -u mysql -f | grep "1297"

If errors continue, check MySQL error log and NDB cluster logs for additional root cause information. Consider scaling the cluster (adding more data nodes) if resource exhaustion persists.