[HTCondor-users] Collector waits for file .collector_address

[Matthias Schnepf <matthias.schnepf@xxxxxxx>]

Hi all,

we are running on our central node one main collector and some sub collectors :

DAEMON_LISTÂÂÂÂ = $(DAEMON_LIST), COLLECTOR, COLLECTOR02

COLLECTOR_ARGS = -sock collector -local-name COLLECTOR01

COLLECTOR02 = $(COLLECTOR)

COLLECTOR02_ARGS = -f -p 20002 -sock collector2 -local-name COLLECTOR02

COLLECTOR02_LOG = $(LOG)/Collector02Log

COLLECTOR02_ENVIRONMENT = 
"_CONDOR_COLLECTOR_LOG=$(LOG)/Collector02Log 
_condor_COLLECTOR_HOST=$(CONDOR_HOST):20002 
_condor_COLLECTOR_NAME=Collector2 _CONDOR_USE_SHARED_PORT=FALSE 
_condor_CONDOR_VIEW_HOST=$(COLLECTOR_HOST)"

When we start or restart the condor service not all daemons starts. Only the main collector and the condor_master started. The other daemons doesn't start and wait. The log file said:

07/24/17 13:51:16 (pid:10856) (D_ALWAYS) Started DaemonCore process "/usr/sbin/condor_collector", pid and pgroup = 10898

07/24/17 13:51:16 (pid:10856) (D_ALWAYS) Waiting for /var/log/condor/.collector_address to appear.

The file /var/log/condor/.collector_address doesn't exist when we start or restart the condor service. After a touch on /var/log/condor/.collector_address everything works fine and the missing daemons starts regularly.

We saw this issue in version 8.6.1 and also after an update on 8.6.5. Do we have to add some configurations?

Thanks

Matthias