12 April,16 at 11:07 AM
Applies to: All versions of Centrify DirectControl on all platforms
Question:
The following log messages can be seen:
lc2121 runmappers[6072]: /usr/share/centrifydc/mappers/machine/copyfile.pl taking too long, killing
lc2121 runmappers[6072]: /usr/share/centrifydc/mappers/machine/copyfile.pl map refresh: Terminated by signal 15
lc2121 runmappers[6072]: /usr/share/centrifydc/mappers/machine/copyfilev2.pl taking too long, killing
lc2121 runmappers[6072]: /usr/share/centrifydc/mappers/machine/copyfilev2.pl map refresh: Terminated by signal 15
lc2121 P: [3646]: ERROR cdcwatch detected adclient is not running properly (ping took 71 seconds)
lc2121 adclient[12780]: WARN com.centrify.logger CircularBuffer.init() - initialized and enabled=true
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/BackwardCompatibility.pl taking too long, killing
lc2121 adclient[13140]: ERROR <bg:gp-init> gp.monitored Group Policy: Timeout waiting for /usr/share/centrifydc/mappers/runmappers to complete
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/copyfile.pl taking too long, killing
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/copyfile.pl map boot: Terminated by signal 15
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/copyfilev2.pl taking too long, killing
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/CrontabEntries.pl taking too long, killing
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/issue.pl taking too long, killing
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/nssoverrides.pl taking too long, killing
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/nssoverrides.pl map boot: Terminated by signal 15
lc2121 cdcwatch[13164]: ERROR cdcwatch detected adclient is not running properly (ping took 58 seconds)
lc2121 runmappers[14849]: /usr/share/centrifydc/mappers/machine/syncpass.pl taking too long, killing
lc2121 adclient[20196]: WARN com.centrify.logger CircularBuffer.init() - initialized and enabled=true
lc2121 cdcwatch[22767]: ERROR cdcwatch detected adclient is not running properly (ping took 33 seconds)
lc2121 adclient[2592]: WARN com.centrify.logger CircularBuffer.init() - initalized and enabled=true
Is there any reason for this?
Answer:
The log above shows that runmappers launches the individual GP scripts serially - it launches one, waits for it to end, then launches the next one.
Run time of different GP varies, even the same script varies a lot.
The systems claim they need more time for its GP to process. The issue is not the script itself, but rather the load of the machine at the time. By default each mapper is allowed 30 seconds to complete and all mapper scripts should be completed in 4 minutes (240 seconds) before adclient kills the scripts.
This is configurable in the Centrify configuration file, /etc/centrifydc/centrifydc.conf:
Open the configuration file and change the following lines:
gp.mappers.timeout: 30
gp.mappers.timeout.all: 240
to:
gp.mappers.timeout: 60
gp.mappers.timeout.all: 600
gp.mappers.timeout - This governs how long adclient is going to wait for runmappers before abandoning the attempt to wait for it to end.
gp.mappers.timeout.all - This governs how long runmappers is going to wait for each perl script that it launched. 600 seconds=10 minutes.
Save the file then Run:
# adreload
# adgpupdate