pclient eats 100% CPU when nothing is being processed

[Sebastian Jester]

Thought this had been noted before, but couldn't find a ticket: pclient uses 100% CPU when no processing is happening (as far as I can tell). See clipped top output below for two pclient processes that don't run any analysis commands but use 100% CPU. I believe this happens on every CPU that's been "controller host add"-ed.

This is not a showstopper, but somewhere between a nuisance if a host is locked up with idle pclient processes and a waste of electricity and hence operations funds.

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                                                                                                  
26476 kschmidt  25   0 66744 1032  980 R  100  0.0  27041:00 pclient                                                                                                   
29374 kschmidt  25   0 65540 1032  980 R  100  0.0  27040:25 pclient
20994 jester    15   0  119m  36m 1340 S    1  0.5 169:03.00 Xvnc                                                                                                      
 2658 jester    15   0  8748 1372  900 R    1  0.0   0:00.03 top                                                                                                       
17343 jester    15   0  789m 185m  10m S    0  2.4 667:38.38 opera                                                                                                     
    1 root      18   0   804   76   44 S    0  0.0   0:06.61 init                                                                                                      
    2 root      RT   0     0    0    0 S    0  0.0   0:04.65 migration/0                                                                                               
    3 root      34  19     0    0    0 S    0  0.0   0:00.86 ksoftirqd/0                                                                                               
    4 root      RT   0     0    0    0 S    0  0.0   0:35.37 migration/1     

[eugene]

I believe this was a problem with missing usleeps in the task thread, since fixed.

our sleeping pclients do not currently show high loads.