YYYY-MM-DD XX:XX:XX,XXX | WARN | pool-1-thread-1 | SerialAggregateTask | System_Public.sql: Task failed due to uncaught exception | org.postgresql.util.PSQLException: An I/O error occurred while sending to the backend. at org.postgresql.core.v3.QueryExecutorImpl.execute(QueryExecutorImpl.java:399) Caused by: java.net.SocketTimeoutException: Read timed outVMware Cloud Director 10.6.1.x
This issue is caused by a communication timeout between the VCD upgrade utility and the PostgreSQL database backend. During the "Refresh database views" task, the database requires more time than the default 90-second timeout to process complex schema changes, resulting in a SocketTimeoutException and an I/O error at the driver level.
To resolve this issue, increase the JDBC socket timeout configuration to allow the database task sufficient time to complete.
Ensure there is no significant network latency and no firewalls or load balancers between the VCD Cells and the Database server that could be dropping long-lived connections.
Create backups of the "global.properties" and "responses.properties" files located in /opt/vmware/vcloud-director/etc/.
On all cells, edit the "global.properties" and "responses.properties" files located in /opt/vmware/vcloud-director/etc/.
Locate the database.jdbcUrl line in both files and increase the socketTimeout value from 90 to 600.
Example (Before): database.jdbcUrl = jdbc:postgresql://<IP_Address>:5432/vcloud?socketTimeout=90&ssl=true
Example (After): database.jdbcUrl = jdbc:postgresql://<IP_Address>:5432/vcloud?socketTimeout=600&ssl=true
Restart the VMware Cloud Director service on all cells to apply the changes:service vmware-vcd restart
Retry Upgrade: Re-run the VCD database upgrade utility:/opt/vmware/vcloud-director/bin/upgrade