Core Dumps in Oracle home/dbs directory after oracle upgrade

[Former Member]

Dear Experts,

Recently we have upgraded our oracle 10.2.0.4 to 11.2.0.4 and after that core dumps are getting generated in /oracle_home/dbs directory.

We could observe that in DBACOCKPIT SGA available memory is showing as zero. IS it the reason for core dumps?

Please suggest how can we analyse core dumps and resolve that.

Regards,

Karthik

[Former Member]

The value 'available memory' displays information about the amount of SGA memory available for future dynamic SGA resize operations (read from gv$sga_dynamic_free_memory. Unlikely this has relation to generated core dumps. What OS you are using?

[Former Member]

We are using AIX 6.1 TL7

What we can do to avoid those core dumps

[Former Member]

You can use dbx. Some information you can get from errpt.

[Former Member]

Can you please explain what  needs to be done.

[Former Member]

We are getting this error in trace file...

Dump of memory from 0x1100047A4 to 0x1100047A8

1100047A0          10470170                        [.G.p]

ub4 kgcepscinitdone_ [1100047A8, 1100047AC) = 00000000

ub4* kgcepgcryptenginemat_ [1100047B0, 1100047B8) = 00000000 00000000

ub4 kgcepgcryptengine_ [1100047B8, 1100047BC) = 00000000

ub4 kgcedbgmode_ [1100047BC, 1100047C0) = 00000000

ub4 kgftrclvl_ [1100047C0, 1100047C4) = 00000000

kgefa * kgfefac_ [1100047C8, 1100047D0) = 00000000 00000000

void * kgfspare_ [1100047D0, 1100047D8) = 00000000 00000000

ub4 kgfmInited_ [1100047D8, 1100047DC) = 00000000

struct kgfmctx * kgfmpgctx_ [1100047E0, 1100047E8) = 00000000 00000000

ssexhd: crashing the process...

Please check and suggest.

[Former Member]

Sorry for misunderstanding. What type of core dumps you mean? These tools are not relevant for dumps generated by Oracle.

[JamesZ]

Hi Karthik,

Please attach the oracle alert log.

Btw from the mentioned location, the dump looks like oracle core dump.

Best regards,

James

[Former Member]

I cannot attach alert log since it is big file.

Yes it is coracle core dump.

Regards,

Karthik

[JamesZ]

Hi Karthik,

If so, please just list some of the alert log, please make sure the content of the timestamps include the time when the core dump generated if possible.

Best regards,
James

[Former Member]

We are not seeing any error kind of entries in alert log file... only trace files the above error its showing...

Dump of memory from 0x1100047A4 to 0x1100047A8

1100047A0          10470170                        [.G.p]

ub4 kgcepscinitdone_ [1100047A8, 1100047AC) = 00000000

ub4* kgcepgcryptenginemat_ [1100047B0, 1100047B8) = 00000000 00000000

ub4 kgcepgcryptengine_ [1100047B8, 1100047BC) = 00000000

ub4 kgcedbgmode_ [1100047BC, 1100047C0) = 00000000

ub4 kgftrclvl_ [1100047C0, 1100047C4) = 00000000

kgefa * kgfefac_ [1100047C8, 1100047D0) = 00000000 00000000

void * kgfspare_ [1100047D0, 1100047D8) = 00000000 00000000

ub4 kgfmInited_ [1100047D8, 1100047DC) = 00000000

struct kgfmctx * kgfmpgctx_ [1100047E0, 1100047E8) = 00000000 00000000

ssexhd: crashing the process...

can you please tell me what is the highlighted thing in that.... is it any memory block or what it is... why because in all core dumps its the one repeating

[ashish_vikas]

Do you see any errors in alert_SID.log file for duration when core dumps got generated ?

Regards,

Ashish

[Former Member]

There are no errors reported in alertSID log related to core dumps... but I can see in one trace file some memory related process is crashed and core dump generated

regards

karthik