db2diag.log error

[tdavidge]

I've just noticed the following errors in my db2diag.log file:

2003-07-18-01.29.18.203000-240 E3301H585 LEVEL: Severe (OS)
PID : 1532 TID : 2944 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : DBNAME
APPHDL : 0-958 APPID: *LOCAL.DB2.030718052918
FUNCTION: DB2 UDB, oper system services, getPrivateChunksFromOs,
probe:100
CALLED : OS, -, VirtualAlloc
OSERR : 8 "Not enough storage is available to process this command."
MESSAGE : Private memory and/or virtual address space exhausted
DATA #1 : Requested size, PD_TYPE_MEM_REQUESTED_SIZE, 4 bytes
43008000

2003-07-18-01.29.18.265000-240 E3888H585 LEVEL: Severe (OS)
PID : 1532 TID : 2944 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : DBNAME
APPHDL : 0-958 APPID: *LOCAL.DB2.030718052918
FUNCTION: DB2 UDB, oper system services, getPrivateChunksFromOs,
probe:100
CALLED : OS, -, VirtualAlloc
OSERR : 8 "Not enough storage is available to process this command."
MESSAGE : Private memory and/or virtual address space exhausted
DATA #1 : Requested size, PD_TYPE_MEM_REQUESTED_SIZE, 4 bytes
43008000

-----------------

I wasn't doing anything specific at the time of the errors [forget the
dates as I was backtesting some apps], I had just finished a database
backup and a db2move, but that had finished around 20 mins before the
two errors occurred.

Can anyone please shed some light on this as I would hate to run into
problems at a more critical juncture.

Win Server 2000, 4 gb mem, DB2 8.2.2 ESE Fp9 applied. With all my
present settings my db2syscs.exe process is hovering around 1725mb of
memory utilized. I have not seen it go beyond that no matter what I do.
I had some earlier problems with over allocating the BP's and crashing
when it blew through the 2gb mark .. Ugh ! Lesson learned there !

Instance memory settings:

MIN_PRIV_MEM = 512
PRIV_MEM_THRESH = 32767

There are no errors in the server OS event viewer so I am at a loss as
to what caused the error and why. Never had the erros before and
haven't changed any instance or database configuration params in ages.

Anyway, any help would be greatly appreciated.

T

[Serge Rielau]

td******@hotmail.com wrote:

I've just noticed the following errors in my db2diag.log file:

2003-07-18-01.29.18.203000-240 E3301H585 LEVEL: Severe (OS)
PID : 1532 TID : 2944 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : DBNAME
APPHDL : 0-958 APPID: *LOCAL.DB2.030718052918
FUNCTION: DB2 UDB, oper system services, getPrivateChunksFromOs,
probe:100
CALLED : OS, -, VirtualAlloc
OSERR : 8 "Not enough storage is available to process this command."
MESSAGE : Private memory and/or virtual address space exhausted
DATA #1 : Requested size, PD_TYPE_MEM_REQUESTED_SIZE, 4 bytes
43008000

2003-07-18-01.29.18.265000-240 E3888H585 LEVEL: Severe (OS)
PID : 1532 TID : 2944 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : DBNAME
APPHDL : 0-958 APPID: *LOCAL.DB2.030718052918
FUNCTION: DB2 UDB, oper system services, getPrivateChunksFromOs,
probe:100
CALLED : OS, -, VirtualAlloc
OSERR : 8 "Not enough storage is available to process this command."
MESSAGE : Private memory and/or virtual address space exhausted
DATA #1 : Requested size, PD_TYPE_MEM_REQUESTED_SIZE, 4 bytes
43008000

-----------------

I wasn't doing anything specific at the time of the errors [forget the
dates as I was backtesting some apps], I had just finished a database
backup and a db2move, but that had finished around 20 mins before the
two errors occurred.

Can anyone please shed some light on this as I would hate to run into
problems at a more critical juncture.

Win Server 2000, 4 gb mem, DB2 8.2.2 ESE Fp9 applied. With all my
present settings my db2syscs.exe process is hovering around 1725mb of
memory utilized. I have not seen it go beyond that no matter what I do.
I had some earlier problems with over allocating the BP's and crashing
when it blew through the 2gb mark .. Ugh ! Lesson learned there !

Instance memory settings:

MIN_PRIV_MEM = 512
PRIV_MEM_THRESH = 32767

There are no errors in the server OS event viewer so I am at a loss as
to what caused the error and why. Never had the erros before and
haven't changed any instance or database configuration params in ages.

Anyway, any help would be greatly appreciated.

T

Not knowing Windows in that depth, but one often overlooked problem in
*nix are the memory quotas for the userid under which DB2 runs...
Just a poke in the dark.

Cheers
Serge
--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

[Matt Emmerton]

<td******@hotmail.com> wrote in message
news:11**********************@g47g2000cwa.googlegr oups.com...

I've just noticed the following errors in my db2diag.log file:

2003-07-18-01.29.18.203000-240 E3301H585 LEVEL: Severe (OS)
PID : 1532 TID : 2944 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : DBNAME
APPHDL : 0-958 APPID: *LOCAL.DB2.030718052918
FUNCTION: DB2 UDB, oper system services, getPrivateChunksFromOs,
probe:100
CALLED : OS, -, VirtualAlloc
OSERR : 8 "Not enough storage is available to process this command."
MESSAGE : Private memory and/or virtual address space exhausted
DATA #1 : Requested size, PD_TYPE_MEM_REQUESTED_SIZE, 4 bytes
43008000

2003-07-18-01.29.18.265000-240 E3888H585 LEVEL: Severe (OS)
PID : 1532 TID : 2944 PROC : db2syscs.exe
INSTANCE: DB2 NODE : 000 DB : DBNAME
APPHDL : 0-958 APPID: *LOCAL.DB2.030718052918
FUNCTION: DB2 UDB, oper system services, getPrivateChunksFromOs,
probe:100
CALLED : OS, -, VirtualAlloc
OSERR : 8 "Not enough storage is available to process this command."
MESSAGE : Private memory and/or virtual address space exhausted
DATA #1 : Requested size, PD_TYPE_MEM_REQUESTED_SIZE, 4 bytes
43008000

-----------------

I wasn't doing anything specific at the time of the errors [forget the
dates as I was backtesting some apps], I had just finished a database
backup and a db2move, but that had finished around 20 mins before the
two errors occurred.

Can anyone please shed some light on this as I would hate to run into
problems at a more critical juncture.

Win Server 2000, 4 gb mem, DB2 8.2.2 ESE Fp9 applied. With all my
present settings my db2syscs.exe process is hovering around 1725mb of
memory utilized. I have not seen it go beyond that no matter what I do.
I had some earlier problems with over allocating the BP's and crashing
when it blew through the 2gb mark .. Ugh ! Lesson learned there !

What you're seeing is DB2 attempting to allocate more memory than your
system will allow.
FP9 handles some instances of this error condition more gracefully than
others -- especially when allocating sort memory or dbheap memory. As you
noted, exceeding system limits with bufferpool sizes at startup is not a
pretty picture.

In order to know what is causing this we'd need more specific correlation
between what you were doing at the time of the error. (ie, db2 operations
such as backup/move, specific queries, etc.)

--
Matt Emmerton

[tdavidge]

Matt Emmerton wrote:

What you're seeing is DB2 attempting to allocate more memory than your
system will allow.
FP9 handles some instances of this error condition more gracefully than
others -- especially when allocating sort memory or dbheap memory. As you
noted, exceeding system limits with bufferpool sizes at startup is not a
pretty picture.

In order to know what is causing this we'd need more specific correlation
between what you were doing at the time of the error. (ie, db2 operations
such as backup/move, specific queries, etc.)

--
Matt Emmerton

Matt, to be honest I wasn't doing anything at the time, I had
successfuly finished a backup and db2move process. The process finished
30 mins prior to the error. It re-occurred over and over every few
hours until I finally stopped and started the instance.

I haven't stopped and started the instance for approx 3 weeks so it
appears to have taken this long before it blew some parameter
somewhere.

Any ideas ?

[Mark A]

<td******@hotmail.com> wrote in message
news:11**********************@o13g2000cwo.googlegr oups.com...

Matt, to be honest I wasn't doing anything at the time, I had
successfuly finished a backup and db2move process. The process finished
30 mins prior to the error. It re-occurred over and over every few
hours until I finally stopped and started the instance.

I haven't stopped and started the instance for approx 3 weeks so it
appears to have taken this long before it blew some parameter
somewhere.

Any ideas ?

Memory leaks should not be ruled out, especially on Windows.

[Matt Emmerton]

"Mark A" <no****@nowhere.com> wrote in message
news:R9********************@comcast.com...

<td******@hotmail.com> wrote in message
news:11**********************@o13g2000cwo.googlegr oups.com...

Matt, to be honest I wasn't doing anything at the time, I had
successfuly finished a backup and db2move process. The process finished
30 mins prior to the error. It re-occurred over and over every few
hours until I finally stopped and started the instance.

I haven't stopped and started the instance for approx 3 weeks so it
appears to have taken this long before it blew some parameter
somewhere.

Any ideas ?

Memory leaks should not be ruled out, especially on Windows.

A good way to track this (and will help convince the DB2 service folks that
you really have a memory leak problem) is to take database snapshots [ db2
get snapshot for database <dbname> ] on a regular basis -- on your system,
every hour or two would probably be sufficient, especially since it takes 3
weeks for you to hit a critical error condition.

A database snapshot will show the size of every memory pool that has been
allocated, including the current allocation level and high-water-mark
levels. If this really is a memory leak, then we should see the current
allocation in some pools increasing over time. My guess is that the problem
memory pool will be the "database heap" (aka dbheap or dbh).

--
Matt Emmerton