Symptom
- An out of memory event happened in the HANA database and an oom trace file was created containing the information below:
[MEMORY_OOM] Information about current out of memory situation: (<timestamp>)
OUT OF MEMORY occurred.
Host: <hosrname>
Executable: <servicename>
PID: <pid>
Failed to allocate <memoryvalue>
Allocation failure type: OPERATING_SYSTEM_LIMIT
- The oom trace file also contains the messages below:
Could not return Xb to operating system.
This is a configuration problem of your operating system:
Please increase /proc/sys/vm/max_map_count
At most one OOM dump is written in approximately Xs.
/proc/sys/vm/max_map_count:
65530 <--- this is far too low
- You also see the following messages in the traces of the HANA service that reported the out of memory event:
<timestamp> e Memory Precharge.cpp(00162) : memAllocSystemPages failed with rc=12, 12 (Cannot allocate memory)
- You could also get an error similar to:
SQL submitted to ODBC data source <HANA SID> resulted in error <[SAP AG][LIBODBCHDB SO][HDBODBC] General error;2048 column
store error: search table error: [9] Memory allocation failed;exception 1000002: Allocation failed ; $size$=25315874691;
$name$=Pool/JoinEvaluator; $type$=pool; $inuse_count$=16; $allocated_size$=27891692375; $alignment$=8;
$failure_type$=OPERATING_SYSTEM_LIMIT;
- The OOM dump files:
<service>_<host>.<port>.rtedump.<timestamp>.oom.trc
<service>_<host>.<port>.rtedump.<timestamp>.oom_memory_release.trc
Show that the sum of U (used memory) is signifcantly smaller than the total allocation limit under AB (allocated bytes) for that specific service. For example the indexserver nodes of a HANA system here :
[13] PID=149687 (state=S), SId=121674146, compactors active, alive, process name: hdbindexserverAB=72.48gb (77832671232b), UA=0b, U=8.61gb (9252275735b), FSL=1.12gb (1207959552b), FMC=512mb (536870912b), SMAS=101.21gb (108682178560b), SMDS=28.56gb (30676598784b), PAL=2.21tb (2439313883136b), TPAL=72.59gb (77949612032b), W=100
[15] PID=596823 (state=R), SId=121674148, compactors active, alive, process name: hdbindexserver
AB=1.10tb (1220820975616b), UA=0b, U=574.47gb (616837124628b), FSL=0b, FMC=0b, SMAS=10.27tb (11296726528000b), SMDS=9.16tb (10075732643840b), PAL=2.21tb (2439313883136b), TPAL=2.18tb (2402562605056b), W=100
- BW DSO activation fails with below errors:
"Message text transaction rolled back by an internal error: fail ed to activate dso with persisted changelog: [410 02] column store error and "Message text Serious error occurred: column store error: search table error: [2999] general error (no further information available) at ptim"
Read more...
Environment
SAP HANA, platform edition
Product
Keywords
SAP HANA, platform edition, sap_hana, database, db, saphana, oom, out-of-memory, out-memory, out memory, operating_system_limit, rc=12, hdbindexserver, hdbscriptserver, hdbnameserver, hdbcompileserver, allocationlimit, global_allocation_limit, memorymanager, max_map_count , KBA , HAN-DB-ENG , SAP HANA DB Engines , Problem
About this page
This is a preview of a SAP Knowledge Base Article. Click more to access the full version on SAP for Me (Login required).Search for additional results
Visit SAP Support Portal's SAP Notes and KBA Search.
SAP Knowledge Base Article - Preview