[one-users] scheduler deploy problem

[antoni artigues]

Hello

I have 4 hosts registered:

$ onehost list
  ID NAME              CLUSTER  RVM   TCPU   FCPU   ACPU    TMEM    FMEM
STAT
   0 oscarnode4        default    1    400    400    400    7.8G      0K
on
   1 oscarnode5        default    1    400    400    400    7.8G      0K
on
   2 oscarnode6        default    3    400    400    400    7.8G      0K
on
   3 oscarnode7        default    4    400    400    400    7.8G      0K
on

Each one With 7.8G of RAM

I create 9 virtual machines. Each of one with 1 cpu and 1792MB of ram:
------------------------------
NAME="ubuntuVM"
VCPU = 1
MEMORY = 1792
OS = [ ARCH = "i686" ]
DISK=[IMAGE="vmimage",
      TARGET="sda",
      BUS="scsi"]
NIC=[NETWORK="vmware LAN bridged"]
------------------------------

There are 4 running machines on "oscarnode7" host.

When I create more virtual machines the scheduler sends it to the
"oscarnode7". But it doen't start because there is no more available RAM
on this host.

If I create more machines, always the scheduler sends it to the same
host!!!

Why the scheduler are not asigning another host? For example,
"oscarnode4" is only running one virtual machine


This is the virtual machines distribution after creating a new virtual
machine:

 onevm list
   ID     USER     NAME STAT CPU     MEM        HOSTNAME        TIME
   38 oneadmin ubuntuVM runn   0      0K      oscarnode7 02 19:04:46
   39 oneadmin ubuntuVM runn   0      0K      oscarnode7 02 18:42:19
   40 oneadmin ubuntuVM runn   0      0K      oscarnode7 02 18:24:08
   42 oneadmin ubuntuVM runn   0      0K      oscarnode6 02 18:07:58
   43 oneadmin ubuntuVM runn   0      0K      oscarnode5 02 18:07:51
   44 oneadmin ubuntuVM runn   0      0K      oscarnode4 02 18:07:43
   46 oneadmin ubuntuVM runn   0      0K      oscarnode6 02 18:06:43
   51 oneadmin ubuntuVM runn   0      0K      oscarnode6 02 17:58:05
   53 oneadmin ubuntuVM prol   0      0K      oscarnode7 00 00:00:43

Thanks in advance

Antoni Artigues