DIA8533C The system memory limit was reached

[jane]

We are getting this error during a large large load.

2006-06-03-02.58.31.688266 Instance:sieinst Node:000
PID:1286274(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4CE.0671B3070116
database utilities DIAG_ERROR Probe:0 Database:SRMW

LOADID: 1286274.2006-06-03-02.58.31.581068.0 (4;125)

Error loading partition. 0, fffffffffffff43d, Detected in file:
sqluvtld.C,line 5060

2006-06-03-02.58.31.690633 Instance:sieinst Node:000
PID:1286274(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4CE.0671B3070116
database utilities sqluLoadPartition Probe:30 Database:SRMW

Load Error: Error loading partition.
2006-06-03-02.58.31.717130 Instance:sieinst Node:000

PID:1286274(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4CE.0671B3070116
oper system services sqlormdir Probe:100 Database:SRMW

errno:
0x0FFFFFFFFFFF0CA0 : 0x00000011 ....

PID:1286274 TID:1 Node:000 Title: Path/Filename
/srmw_prod/sieinst/NODE0000/SQL00001/load/DB200004.PID/DB20007D.OID

2006-06-03-02.59.15.275385 Instance:sieinst Node:000
PID:1146946(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4EC.017313070159
database utilities sqluAllocTBufs Probe:100 Database:SRMW

DIA8533C The system memory limit was reached.

PID:1146946 TID:1 Node:000 Title: commacb
Dump File:/homeudb/sieinst/sqllib/db2dump/11469461.000

We created the following buffer pools

create bufferpool bp_data1_32K size 100000 pagesize 32K
create bufferpool bp_data2_32K size 100000 pagesize 32K
create bufferpool bp_temp1_32K size 30000 pagesize 32K
Is it running out of bufferpool memory or is it running out of AIX
memory?

Thanks,
-Jane

[m0002a]

jane wrote:

We are getting this error during a large large load.

2006-06-03-02.58.31.688266 Instance:sieinst Node:000
PID:1286274(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4CE.0671B3070116
database utilities DIAG_ERROR Probe:0 Database:SRMW

LOADID: 1286274.2006-06-03-02.58.31.581068.0 (4;125)

Error loading partition. 0, fffffffffffff43d, Detected in file:
sqluvtld.C,line 5060

2006-06-03-02.58.31.690633 Instance:sieinst Node:000
PID:1286274(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4CE.0671B3070116
database utilities sqluLoadPartition Probe:30 Database:SRMW

Load Error: Error loading partition.
2006-06-03-02.58.31.717130 Instance:sieinst Node:000

PID:1286274(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4CE.0671B3070116
oper system services sqlormdir Probe:100 Database:SRMW

errno:
0x0FFFFFFFFFFF0CA0 : 0x00000011 ....

PID:1286274 TID:1 Node:000 Title: Path/Filename
/srmw_prod/sieinst/NODE0000/SQL00001/load/DB200004.PID/DB20007D.OID

2006-06-03-02.59.15.275385 Instance:sieinst Node:000
PID:1146946(db2agent (SRMW) 0) TID:1
Appid:GAFE423C.P4EC.017313070159
database utilities sqluAllocTBufs Probe:100 Database:SRMW

DIA8533C The system memory limit was reached.

PID:1146946 TID:1 Node:000 Title: commacb
Dump File:/homeudb/sieinst/sqllib/db2dump/11469461.000

We created the following buffer pools

create bufferpool bp_data1_32K size 100000 pagesize 32K
create bufferpool bp_data2_32K size 100000 pagesize 32K
create bufferpool bp_temp1_32K size 30000 pagesize 32K
Is it running out of bufferpool memory or is it running out of AIX
memory?

Thanks,
-Jane

You have allocated over 10GB in your bufferpools. If you are using a
32-bit instance, then the maximum for AIX is about 1.75GB. Please post
the output of db2level to show the release and instance type (32 or 64
bit).

[m0002a]

[e-mail] wrote:

You have allocated over 10GB in your bufferpools. If you are using a
32-bit instance, then the maximum for AIX is about 1.75GB. Please post
the output of db2level to show the release and instance type (32 or 64
bit).

Correction. You have allocated over 7GB for your bufferpools.

[jane]

Ours is a 64-bit instance. Is the buffer pool too high? We are
loading large amounts of data into the database using Informatica.

$ db2level
DB21085I Instance "db2inst1" uses "64" bits and DB2 code release
"SQL08015"
with level identifier "02060106".
Informational tokens are "DB2 v8.1.1.50", "siebel_11055",
"U496793_11055", and
FixPak "5".
Product is installed at "/usr/opt/db2_08_01".
Thanks for your help.

[jane]

Here is the setting for unlimit.

ulimit -a

time(seconds) unlimited
file(blocks) unlimited
data(kbytes) 245760 ======> should we change it
unlimited?
stack(kbytes) 16384
memory(kbytes) unlimited
coredump(blocks) unlimited
nofiles(descriptors) 2000

[Mark A]

"jane" <ja**********@sympatico.ca> wrote in message
news:11**********************@h76g2000cwa.googlegr oups.com...

Here is the setting for unlimit.

ulimit -a

time(seconds) unlimited
file(blocks) unlimited
data(kbytes) 245760 ======> should we change it
unlimited?
stack(kbytes) 16384
memory(kbytes) unlimited
coredump(blocks) unlimited
nofiles(descriptors) 2000

Please run db2level and post results.

[jane]

Here it is again ...

$ db2level
DB21085I Instance "db2inst1" uses "64" bits and DB2 code release
"SQL08015"
with level identifier "02060106".
Informational tokens are "DB2 v8.1.1.50", "siebel_11055",
"U496793_11055", and
FixPak "5".
Product is installed at "/usr/opt/db2_08_01".

[Liam Finnie]

jane wrote:

Here it is again ...

$ db2level
DB21085I Instance "db2inst1" uses "64" bits and DB2 code release
"SQL08015"
with level identifier "02060106".
Informational tokens are "DB2 v8.1.1.50", "siebel_11055",
"U496793_11055", and
FixPak "5".
Product is installed at "/usr/opt/db2_08_01".

Hi Jane,

Please take a look at your utility heap setting (UTIL_HEAP_SZ database
configuration parameter) - sqluAllocTBufs (the function that mentioned
the system memory limit was reached) tries to allocate from that heap.

Cheers,
Liam.

[jane]

Thanks for the replies. This is what we have for UTIL_HEAP_SZ. We are
doing large loads. This is big enough?

$ db2 get db cfg for dbname |grep -i UTIL_HEAP_SZ
Utilities heap size (4KB) (UTIL_HEAP_SZ) = 10000

Should I increase this to 100000?

Thanks Again

[Mark A]

"jane" <ja**********@sympatico.ca> wrote in message
news:11*********************@j55g2000cwa.googlegro ups.com...

Here it is again ...

$ db2level
DB21085I Instance "db2inst1" uses "64" bits and DB2 code release
"SQL08015"
with level identifier "02060106".
Informational tokens are "DB2 v8.1.1.50", "siebel_11055",
"U496793_11055", and
FixPak "5".
Product is installed at "/usr/opt/db2_08_01".

You have a 64 bit instance, so you should have no problem with 7GB of
bufferpools.

Looks like you have a special version of DB2 for Siebel. FP 5 is quite old,
and it would nice if you could upgrade to FP12 or a more recent fixpack (but
maybe Siebel does not support past FP5?).

If you have an IBM support contract, open a PMR.

[Mark Yudkin]

40MB sure looks small in relation to your other allocations. Try increasing
it until the problem goes away.

"jane" <ja**********@sympatico.ca> wrote in message
news:11**********************@y43g2000cwc.googlegr oups.com...

Thanks for the replies. This is what we have for UTIL_HEAP_SZ. We are
doing large loads. This is big enough?

$ db2 get db cfg for dbname |grep -i UTIL_HEAP_SZ
Utilities heap size (4KB) (UTIL_HEAP_SZ) = 10000

Should I increase this to 100000?

Thanks Again

[Joachim Müller]

Jane,

this could be the solution. Some time ago we had the same problem,
but the message in the db2diag.log was 'the memory and/or user limit was
reached'.
Give it a try. How big are your container files. If they are smaller than 2
GB you have no problem.

You have to change to unlimited and the restart the instance.

regards,
Joachim Müller

"jane" <ja**********@sympatico.ca> schrieb im Newsbeitrag
news:11**********************@h76g2000cwa.googlegr oups.com...

Here is the setting for unlimit.

ulimit -a

time(seconds) unlimited
file(blocks) unlimited
data(kbytes) 245760 ======> should we change it
unlimited?
stack(kbytes) 16384
memory(kbytes) unlimited
coredump(blocks) unlimited
nofiles(descriptors) 2000

[jane]

I have done the following ...

$ ulimit -a
time(seconds) unlimited
file(blocks) unlimited
data(kbytes) unlimited ===> changed to unlimited
stack(kbytes) 16384
memory(kbytes) unlimited
coredump(blocks) unlimited
nofiles(descriptors) 2000

db2 update db cfg for dbname using UTIL_HEAP_SZ 200000

I am hoping this will fix the problem. I'll you know ...

[jane]

After making the changes we are getting the following error.

[IBM][CLI Driver][DB2/AIX64] SQL3011C There is not enough storage to
process the command.
sqlstate = HY000

should I increase the UTIL_HEAP_SZ to 200000? Or should I open a PMR?

Thanks,
-Jane

[Mark Yudkin]

Probably best to increase memory. Alternatively, shrink the LOAD buffers.

I doubt if a PMR will get you much more than the recommendations you're
being given here. But it probably can't hurt to open one either.

"jane" <ja**********@sympatico.ca> wrote in message
news:11*********************@i40g2000cwc.googlegro ups.com...

After making the changes we are getting the following error.

[IBM][CLI Driver][DB2/AIX64] SQL3011C There is not enough storage to
process the command.
sqlstate = HY000

should I increase the UTIL_HEAP_SZ to 200000? Or should I open a PMR?

Thanks,
-Jane

[jane]

This is the solution to this problem:

1)As root changed the maxclient% from 80 to 20. On AIX 5 the default
is 80.
# vmo -p -o maxclient%=20
Setting maxclient% to 20 in nextboot file Setting maxclient% to 20

2)Changed data and stack to unlimited for instance owner and root.
$ ulimit -a
time(seconds) unlimited
file(blocks) unlimited
data(kbytes) unlimited -> changing to unlimited
stack(kbytes) unlimited -> changing to unlimited
memory(kbytes) unlimited
coredump(blocks) unlimited
nofiles(descriptors) 2000

These changes are required for 64 bit environments.

Thanks for every one's input.