Skip to Content
0
Former Member
Jan 25, 2011 at 09:12 AM

ERROR => User xxx reached avail. heapsize

1406 Views

Hello guys

I have sap SAP ECC 6.0 (kernel 221).

During the night more than one job is executed . I noticed that one of the jobs, which ends before 15.000 sec, reached 46.000 sec and still runing

From st11, i notice :

Mon Jan 24 20:22:03 2011

  • ERROR => User xxx reached avail. heapsize = 3815 MB: see ST22 [abstor.c 1779]

  • ERROR => Increase abap/heap_area_total [abstor.c 1780]

...............

  • ERROR => heap memory WP47: 0 MB [abstor.c 1784]

  • ERROR => heap memory WP48: 194 MB [abstor.c 1784]

  • ERROR => heap memory WP49: 286 MB [abstor.c 1784]

  • ERROR => heap memory WP50: 3071 MB [abstor.c 1784]

  • ERROR => heap memory WP51: 0 MB [abstor.c 1784]

  • ERROR => heap memory WP52: 262 MB [abstor.c 1784]

  • ERROR => heap memory WP53: 0 MB [abstor.c 1784]

................

where , job in question was WP52.

Because we saw "ERROR => heap memory WP50: 3071 MB" I read and log for WP50 , and at that time :

Mon Jan 24 20:22:02 2011

  • ERROR => User YYY reached avail. heapsize = 3815 MB: see ST22 [abstor.c 1779]

  • ERROR => Increase abap/heap_area_total [abstor.c 1780]

So far as I can tell yyy user spent much of heapsize sia , and the job for xxx has not found the free memory .

At this time on st02:

SAP Memory Curr.Use % CurUse[KB] MaxUse[KB] In Mem[KB] OnDisk[KB] SAPCurCach HitRatio %

Roll area 6,52 78.630 160.920 302.080 904.192 IDs 98,38

Page area 12,15 248.779 509.912 655.360 1.392.640 Statement 95,00

Extended memory 19,64 10.457.088 25.223.168 53.243.904 0 0,00

Heap memory 268.835 3.906.915 0 0 0,00

I have memory parameters :

ztta/roll_area 7500800 Byte

ztta/roll_first 1024 Byte

ztta/short_area 4000000 Byte

rdisp/ROLL_SHM 37760 8 kB

rdisp/PG_SHM 81920 8 kB

rdisp/PG_LOCAL 150 8 kB

em/initial_size_MB 52000 MB

em/blocksize_KB 4096 kB

em/address_space_MB 4096 MB

ztta/roll_extension 6500000000 Byte

abap/heap_area_dia 4000683008 Byte

abap/heap_area_nondia 4000683008 Byte

abap/heap_area_total 4000683008 Byte

abap/heaplimit 40894464 Byte

At os (linux sles 10 sp2) level I have

> free -g

total used free shared buffers cached

Mem: 62 46 15 0 1 38

-/+ buffers/cache: 7 54

Swap: 20 0 20

I asked a abaper to look a little on abap code in frogram who allocate 3071 MB from heap area but, meanwhile I want to increase this parameter:

abap/heap_area_dia 6500000000 Byte

abap/heap_area_nondia 6500000000 Byte

abap/heap_area_total 6500000000 Byte

Now I have 2 questions :

1) is okay to to increase abap/heap_area* to a value so high ? or other parameters should increase to avoid such errors ?

and if it is well.. should be increased and abap/heaplimit ?

2) Job that still runs , is hooked and should be stopped? or pull hopes it would end eventually (I can help with something on it)?