[HTCondor-users] job memory requirements and free memory vs cached

[Michael Paterson <mhp@xxxxxxx>]

Hello,

I'm trying to run 4 single core jobs on a 4cpu box with partitionable 
slots and ~7500m memory, the jobs have 1500m mem set for the memory 
requirement.

Sometimes all 4 slots will start up on a machine, but others only get 3.

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+ COMMAND
 7804 slot03    30  10 1381m 787m  26m R 100.0 10.5  63:25.24 basf2
10279 slot02    30  10 1347m 793m  64m R 100.0 10.6  41:47.69 basf2
 6322 slot01    30  10 1546m 891m  21m R 98.4 11.9 386:24.91 basf2

# free -m
             total            used       free     shared buffers     cached
Mem:          7514       6824        689          0         69     3320
-/+ buffers/cache:       3434       4080
Swap:        16383         10      16373


Is the ~3G in cache preventing the 4th job from setting a slot? If so is 
there a way to limit the cachesize or force the machine to give it up so 
condor sees enough free memory to start another job?

I've seen a few references to vm.pagecache but attempting to set it 
gives an error:
# sysctl -w vm.pagecache="20"
error: "vm.pagecache" is an unknown key

the machine is running Scientific Linux release 6.6 (Carbon) and 
vmlinuz-2.6.32-573


thank you