perfSONAR User Q&A and Other Discussion

[Sean Crosby <>]

Hi all,

Today I reinstalled both of my perfSonar boxes to Centos 7/PS 4.0 using the full CD (not netinstall). Everything worked out of the box, except for the creation of the pscheduler user and database.

After install and reboot, I had to do this:

# sudo -u postgres bash

$ createuser pscheduler

$ psql -f /usr/share/pscheduler/database-build-super.sql

$ psql -f /usr/share/pscheduler/database-build.sql

Otherwise I got these errors in the journal

Apr 19 15:03:47 ps-latency journal: safe_run/ticker ERROR    Exception: OperationalError: FATAL:  role "pscheduler" does not exist#012#012Traceback (most recent call last):#012  File "/usr/lib/python2.7/site-packages/pscheduler/saferun.py", line 41, in safe_run#012    function()#012  File "/usr/libexec/pscheduler/daemons/ticker", line 156, in <lambda>#012    pscheduler.safe_run(lambda: main_program())#012  File "/usr/libexec/pscheduler/daemons/ticker", line 123, in main_program#012    db = pscheduler.pg_connection(dsn)#012  File "/usr/lib/python2.7/site-packages/pscheduler/db.py", line 35, in pg_connection#012    pg = psycopg2.connect(dsn)#012  File "/usr/lib64/python2.7/site-packages/psycopg2/__init__.py", line 164, in connect#012    conn = _connect(dsn, connection_factory=connection_factory, async=async)#012OperationalError: FATAL:  role "pscheduler" does not exist

Apr 19 15:03:47 ps-latency journal: safe_run/ticker ERROR    Waiting 8.25 seconds before restarting

Apr 19 15:03:56 ps-latency journal: safe_run/scheduler ERROR    Restarting

Apr 19 15:03:56 ps-latency journal: safe_run/runner ERROR    Restarting

Apr 19 15:03:56 ps-latency journal: safe_run/scheduler ERROR    Program threw an exception after 0:00:00.003040

Apr 19 15:03:56 ps-latency journal: safe_run/runner ERROR    Program threw an exception after 0:00:00.003030

Apr 19 15:03:56 ps-latency journal: safe_run/scheduler ERROR    Exception: OperationalError: FATAL:  database "pscheduler" does not exist#012#012Traceback (most recent call last):#012  File "/usr/lib/python2.7/site-packages/pscheduler/saferun.py", line 41, in safe_run#012    function()#012  File "/usr/libexec/pscheduler/daemons/scheduler", line 660, in <lambda>#012    pscheduler.safe_run(lambda: main_program())#012  File "/usr/libexec/pscheduler/daemons/scheduler", line 527, in main_program#012    pg = pscheduler.pg_connection(dsn)#012  File "/usr/lib/python2.7/site-packages/pscheduler/db.py", line 35, in pg_connection#012    pg = psycopg2.connect(dsn)#012  File "/usr/lib64/python2.7/site-packages/psycopg2/__init__.py", line 164, in connect#012    conn = _connect(dsn, connection_factory=connection_factory, async=async)#012OperationalError: FATAL:  database "pscheduler" does not exist

Apr 19 15:03:56 ps-latency journal: safe_run/scheduler ERROR    Waiting 8.5 seconds before restarting

Not sure if it is just the full CD that has this error, and it has been fixed in the netinstall, but it happened on both of my boxes.

Cheers,

Sean

--

Sean Crosby
Research Computing
ARC Centre of Excellence for Particle Physics at the Terascale
School of Physics | University of Melbourne VIC 3010 Australia