[Bug] gpdemo failed on external FTS build

[congxuebin]

Cloudberry Database version

No response

What happened

20240830:11:34:09:024274 gpstart:i-1ba8pr8n:gpadmin-[WARNING]:-Standby host is unreachable, cannot determine whether the standby is currently acting as the coordinator. Received error: ExecutionError: 'non-zero rc: 1' occurred.  Details: 'ssh -o StrictHostKeyChecking=no -o ServerAliveInterval=60 i-1ba8pr8n ". /usr/local/cloudberry-db-devel/greenplum_path.sh; $GPHOME/bin/pg_controldata /code/gpdb_src/gpAux/gpdemo/datadirs/standby"'  cmd had rc=1 completed=True halted=False
  stdout=''
  stderr='pg_controldata: fatal: could not open file "/code/gpdb_src/gpAux/gpdemo/datadirs/standby/global/pg_control" for reading: No such file or directory
'
20240830:11:34:09:024274 gpstart:i-1ba8pr8n:gpadmin-[WARNING]:-Continue only if you are certain that the standby is not acting as the coordinator.
20240830:11:34:09:024274 gpstart:i-1ba8pr8n:gpadmin-[WARNING]:-Non interactive mode detected. Not starting the cluster. Start the cluster in interactive mode.
20240830:11:34:18:024274 gpstart:i-1ba8pr8n:gpadmin-[INFO]:-User abort requested, Exiting...
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[WARN]:
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[WARN]:-Failed to start CloudberryDB instance; review gpstart output to
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[WARN]:- determine why gpstart failed and reinitialize cluster after resolving
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[WARN]:- issues.  Not all initialization tasks have completed so the cluster
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[WARN]:- should not be used.
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[WARN]:-gpinitsystem will now try to stop the cluster
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[WARN]:
20240830:11:34:18:024825 gpstop:i-1ba8pr8n:gpadmin-[INFO]:-Starting gpstop with args: -a -l /code/gpdb_src/gpAux/gpdemo/datadirs/gpAdminLogs -i -d /code/gpdb_src/gpAux/gpdemo/datadirs/qddir/demoDataDir-1
20240830:11:34:18:024825 gpstop:i-1ba8pr8n:gpadmin-[INFO]:-Gathering information and validating the environment...
20240830:11:34:18:024825 gpstop:i-1ba8pr8n:gpadmin-[ERROR]:-gpstop error: postmaster.pid file does not exist.  is CloudberryDB instance already stopped?
20240830:11:34:18:018462 gpinitsystem:i-1ba8pr8n:gpadmin-[FATAL]:-Error from CloudberryDB instance shutdown, check log files Script Exiting!
========================================
gpinitsystem returned: 1
========================================

What you think should happen instead

No response

How to reproduce

rpm -ivh {external_fts_build}.rpm

gpdemo command would fail

Operating System

centos7

Anything else

No response

Are you willing to submit PR?

• Yes, I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct.

[congxuebin]

Hi @Ray-Eldath Zhaohan, this is the issue we discussed last week. I opened this issue for tracking purpose only.

[congxuebin]

Hi Max, I heard Zhaohan has left the company. Just fyi, we have this issue.

[jiaqizho]

external-fts does not support gpdemo. You can manually write the configuration file if necessary.

[roseduan]

This case only reproduce when export WITH_STANDBY=true

[congxuebin]

Jiaqi, the external-fts build is a release build that we support for testing. I think you mean gpdemo on it will be supported in the future and we keep this issue open as you have lowered the priority.

[jiaqizho]

Jiaqi, the external-fts build is a release build that we support for testing. I think you mean gpdemo on it will be supported in the future and we keep this issue open as you have lowered the priority.

yes