fsmiledream

After set enable_parallel= off,this error not occurred; May the parameter enable_parallel = on have some problem?

But it's a test environment ,only 1 test connect . connections and resqueue is enough ,see pictures below . you can provide info by command: `ulimit -a`

max_parallel_workers_per_gather=64 in my environment. The sql : SELECT A.serial_no, a.visit_start_time, A.visit_end_time, bpi.* FROM (SELECT ov.org_code, ov.person_info_id, ov.serial_no, '1' AS visit_type_code, ov.visit_start_time, ov.visit_end_time FROM visit ov WHERE ov.update_datetime BETWEEN '2024-09-13 13:43:33.180'...

max_parallel_workers_per_gather=64 is also in greenplum .Because the server configuration is enough. There are four physical machines, each with 256GB of memory and 64-core cpu,and ssd disks.

@yihong0618 Thank you very much! Using your solution ,I can also initsystem successfully. I also hope source code can be fixed and updated .

not support multi etcd endpoints，so gpdb doesn't support high availability right now? Is the description of the official website inaccurate? https://cloudberrydb.org/zh/docs/cbdb-op-software-hardware