Good morning,
I am testing a dataguard environment at the moment and I found something in the alert.log, I can't solve:
WARNING: Heavy swapping observed on system in last 5 mins.
Heavy swapping can lead to timeouts, poor performance, and instance eviction.
Errors in file D:\ORACLE\diag\rdbms\test_stby\test_stby\trace\test_stby_dbrm_4964.trc (incident=209923):
ORA-00700: Interne Warnung, Argumente: [kskvmstatact: excessive swapping observed], [], [], [], [], [], [], [], [], [], [], []
Incident details in: D:\ORACLE\diag\rdbms\test_stby\test_stby\incident\incdir_209923\test_stby_dbrm_4964_i209923.trc
2019-05-24T03:19:40.786866+02:00
Dumping diagnostic data in directory=[cdmp_20190524031940], requested by (instance=1, osid=4964 (DBRM)), summary=[incident=209923].
Sometimes I get an additional message in between:
WARNING: Heavy swapping observed on system in last 5 mins.
Heavy swapping can lead to timeouts, poor performance, and instance eviction.
ERROR: Unable to normalize symbol name for the following short stack (at offset 182):
dbgexProcessError()+398<-dbgePostErrorKGE()+2191<-dbkePostKGE_kgsf()+90<-kgeadse()+457<-kgerinv_internal()+86<-kgesoftnmierr()+768<-kskvmstatact()+1602<-kskdbrmtoutact()+259<-ksbcti..1()+348<-ksbabs()+15825<-ksbrdp()
+1556<-opirip()+1539<-opidrv()+960<-sou2o()+112<-opimai_real()+222<-opimai()+231<-BackgroundThreadStart()+702<-00007FFE7BFA84D4<-00007FFE7C25E851
It seems like, that this situation occurs only to the database, which has the standby role at this moment. The dataguard environment consists of one primary db and one standby db (Standalone, no RAC) which run in maxavailibility mode.
I found some solutions online, which suggest to increase the overall memory or to decrease sga_target or pga_aggregate_target or and pga_aggregate_limit.
Even the trace files suggest this.
I reduced some values. But still the same result. The problem occurs not every hour, but a couple of times in 24 hours.
Here the memory configuration:
memory_max_target big integer 0
memory_target big integer 0
pga_aggregate_limit big integer 10848M
pga_aggregate_target big integer 5424M
sga_target big integer 9600M <--- I decreased this value from 17GB for testing
sga_max_size big integer 10016M <--- I decreased this value from 17GB for testing
Here some more information:
total physical memory: 32767 MB
bytes swapped in/out in last 5 mins: 779775 KB
cpu utilization: 1%
SGA_TARGET: 9600 MB
PGA_AGGREGATE_LIMIT: 10848 MB
maximum PGA allocated since startup: 373 MB
total PGA currently allocated: 373 MB
PGA_AGGREGATE_TARGET: 5424 MB
ORA-00700: Interne Warnung, Argumente: [kskvmstatact: excessive swapping observed], [], [], [], [], [], [], [], [], [], [], []
Heavy swapping observed in last 5 mins: |pct of total memory][bytes]
total swpin [ 6.94%][2330311K], total swpout [ 0.00%][0K]
vm stats captured every 30 secs for last 5 mins:
swpin:swpout:
[0.02%][7566K][0.00%][0K]
[0.30%][102901K][0.00%][OK]
[0.27%][91913K][0.00%][0K]
[0.08%][29549K][0.00%][0K]
[5.01%][1683433K][0.00%][0K]
[0.51%][172509K][0.00%][0K]
[0.64%][215786K][0.00%][0K]
[0.02%][8386K][0.00%][0K]
[0.02%][8190K][0.00%][0K]
[0.03%][10073K][0.00%][0K]
2019-05-23T18:18:47.069181+02:00
Incident 204890 created, dump file: D:\ORACL....
The server uses only 12GB of 32GB.
Do you have any idea, what could be the issue? Any suggestions?
Thank you!
Best regards
IronMike