We are currently running about 120 Linux instances on our 1 IFL G5 with 1.5G
storage.

After I added about the 100th Linux, VM began refusing to dispatch some of
them for extremely long periods of time (hours).  If I re-IPL VM and start
all 120 Linux systems up again it runs fine for a couple days, then the
problem starts occurring again.  I Think a VM resource is being over
committed or something, so VM just decides it doesn't have the resource to
run some of them anymore so it puts them on the eligible list, but, never
dispatches them again or at least not until a resource becomes available
again which could be hours later.

INDICATE Shows:

AVGPROC-034% 01
MDC READS-000001/SEC WRITES-000000/SEC HIT RATIO-095%
STORAGE-096% PAGING-0002/SEC STEAL-000%
Q0-00002(00001)                           DORMANT-00012
Q1-00000(00000)           E1-00000(00000)
Q2-00000(00000) EXPAN-001 E2-00000(00000)
Q3-00115(00000) EXPAN-002 E3-00008(00006)

PROC 0000-034%

LIMITED-00000

Here's what RTM shows:

<USERID> %CPU %CP %EM ISEC PAG  WSS  RES   UR PGES SHARE VMSIZE TYP,CHR,STAT

VMSERVU  IDLE .00 .00  .00 .00 1178    2   .0 1178  1500    32M VUC,QDS,IDLE
VMSERVS  IDLE .00 .00  .00 .00 1169    2   .0 1169  1500    32M VUC,QDS,IDLE
OPERSYMP IDLE .00 .00  .00 .00 1193    0   .0 1193   100    32M VUX,DSC,IDLE
EREP     IDLE .00 .00  .00 .00   67    0   .0 1164   100    32M VUX,IAB,IDLE
FTPSERVE IDLE .00 .00  .00 .00   19   21   .0 1589   100    16M VUC,IAB,IDLE
CDM0007  IDLE .00 .00  .00 .00 3934 3335   .0 5373   100    32M VUS,DSC,ELIG
CDM0091  IDLE .00 .00  .00 .00 4455 3764   .0 5163   100    32M VUS,DSC,ELIG
CDM0092  IDLE .00 .00  .00 .00 7064   31   .0 7642   100    32M VUS,DSC,ELIG
CDM0099  IDLE .00 .00  .00 .00 7854   28   .0 7630   100    32M VUS,DSC,ELIG
CDM0104  IDLE .00 .00  .00 .00 7710   32   .0 7671   100    32M VUS,DSC,ELIG
CDM0105  IDLE .00 .00  .00 .00 7958 3725   .0 7634   100    32M VUS,DSC,ELIG
CDM0086  IDLE .00 .00  .00 .00 7880   26   .0 7464   100    32M VUS,DSC,ELIG
IBML01   IDLE .00 .00  .00 .00  16K  16K   .0    1   200    64M VUS,DSC,ELIG


All of the users with a status of ELIG in the list are Linux systems.  There
seems to be plenty of CPU available and paging is light.  Anyone got an idea
what might be causing VM to not dispatch these users?
_____________________________________
Lonny Sivey
System Support Division
OCLC Online Computer Library Center, Inc.
6565 Frantz Rd, Dublin, OH 43017
(614) 764-6013  FAX (614) 718-7200
mailto:[EMAIL PROTECTED]
_____________________________________

Reply via email to