Puppet Agent 3.0.1 intermitently doesn't wake up

[MasterPO]

I have 39 RHEL nodes running the puppet agent and intermitently from 1 to 3 nodes will go unresponsive and require intervention to become active again.

I collected the following for one such instance from this morning:

Hung Puppet agent information:

[root@anmvwms3 ~]# rpm -qa | grep puppet
puppet-3.0.1-1.el5


[root@anmvwms3 ~]# ps -ef | grep puppet
root     12421     1  0 Dec10 ?        00:03:02 /usr/bin/ruby /usr/bin/puppet agent --server=puppet --logdest=/var/log/puppet/puppet.log

[root@anmvwms3 ~]# ps -eo pid,ppid,state,comm,time,pri,size,wchan | grep puppet
12421     1 S puppet          00:03:02  21 43028 -

[root@anmvwms3 ~]# ps -elf | grep 12421
5 S root     12421     1  0  78   0 - 12877 -      Dec10 ?        00:03:02 /usr/bin/ruby /usr/bin/puppet agent --server=puppet --logdest=/var/log/puppet/puppet.log


[root@anmvwms3 ~]# cat /proc/12421/status
Name:   puppet
State:  S (sleeping)
SleepAVG:       78%
Tgid:   12421
Pid:    12421
PPid:   1
TracerPid:      0
Uid:    0       0       0       0
Gid:    0       0       0       0
FDSize: 32
Groups: 0 1 2 3 4 6 10
VmPeak:    52620 kB
VmSize:    51508 kB
VmLck:         0 kB
VmHWM:     36772 kB
VmRSS:     36012 kB
VmData:    42748 kB
VmStk:       280 kB
VmExe:         4 kB
VmLib:      8104 kB
VmPTE:       116 kB
StaBrk: 08cb6000 kB
Brk:    0a404000 kB
StaStk: bfc8b270 kB
Threads:        1
SigQ:   0/81920
SigPnd: 0000000000000000
ShdPnd: 0000000000000000
SigBlk: 0000000000000000
SigIgn: 0000000000000000
SigCgt: 0000000182007e47
CapInh: 0000000000000000
CapPrm: 00000000fffffeff
CapEff: 00000000fffffeff
Cpus_allowed:   00000001
Mems_allowed:   1


Right now, I just restart puppet on the node to get it to resume functioning.

Is this a known issue?

[David Mesler]

Do they ever wake up on their own? I just posted about my issue where every once in a while my agents will sleep for an hour even though they're configured to run every 15 (+ 7.5 splay) minutes.

--david