Changing the Neo4j Backup Client Timeout

Neo4j 4.x, 5.x

Attempting to perform full or incremental backups, might fail due to server unavailability and therefore unresponsiveness, for a prolonged time period. In some cases, the following stacktrace appears at the Client side when executing the neo4j-admin backup command:

Execution of backup failed. Request timed out with no responses after 600000 ms

A more detailed stacktrace is below:

neo4j-admin backup --backup-dir=/var/opt/neo4j/backup --database=neo4j --pagecache=8G --fallback-to-full=true --check-consistency=false --parallel-recovery=false

[INFO] [CAAPM.AgentJPMSEnabler] Starting Agent module configuration for traditional/classpath launch
[INFO] [CAAPM.AgentJPMSEnabler] Finishing up Agent module configuration
INFO [c.n.b.v.BackupOutputMonitor] Start receiving transactions from 123456789
INFO [c.n.b.v.BackupOutputMonitor] Finished, took 10m 4s 352ms
Execution of backup failed. Request timed out with no responses after 600000 ms.

Whilst above indicates underlying issues such as local or network resource bottlenecks, impacting server availability and such underlying issues must be addresses, the following configuration can be altered from its default value of 10m, to change the timeout threshold:

For Neo4j 4.x:

causal_clustering.catch_up_client_inactivity_timeout

For Neo4j 5.x:

dbms.cluster.catchup.client_inactivity_timeout

The configuration title may not immediately seem related to backups, however, it controls the timeout duration of the catchup protocol, which is the underlying protocol of multiple catchup processes, including backups.

Reference: https://neo4j.com/docs/operations-manual/current/configuration/configuration-settings/#config_dbms.cluster.catchup.client_inactivity_timeout