TSV_TNEW_PAGE_ALLOC_FAILED

[Former Member]

Hi Experts,

Daily we are getting the 20-30 memory related dumps for below shopping cart program .

Can someone please check and give recommended values to avoid memory issues .

Program: SAPLBBP_PD OS:Windows NT 64 bit Database: MSSQL

ST22: Memory consumption

Roll.... 2718272

EM...... 2002738624

Heap.... 8000009088

Page.... 0 MM

Used. 9667112304

MM Free. 315216880

Configured memory in system:

ztta/roll_first 1 Byte First amount of roll area used in a dialog WP

ztta/roll_extension 2000000000 Byte Max. extended mem. per session (external mode) ztta/roll_area 3000000 Byte Roll area per workprocess (total)

abap/heap_area_dia 8000000000 Byte Max. heap memory for dialog workprocesses ztta/short_area 3200000 Byte Short area per workprocess

rdisp/ROLL_SHM 32768 8 kB Part of roll file in shared memory

rdisp/PG_SHM 131072 8 kB Part of paging file in shared memory

rdisp/PG_LOCAL 150 8 kB Paging buffer per workprocess

em/initial_size_MB 16384 MB Initial size of extended memory

em/blocksize_KB 4096 kB Size of one extended memory block

em/address_space_MB 4096 MB Address space reserved for ext. mem. (NT only) ztta/roll_extension 2000000000 Byte Max. extended mem. per session (external mode) abap/heap_area_dia 8000000000 Byte Max. heap memory for dialog workprocesses abap/heap_area_nondia 0 Byte Max. heap memory for non-dialog workprocesses abap/heap_area_total 17179869184 Byte Max. usable heap memory abap/heaplimit 40000000 Byte Workprocess restart limit of heap memory abap/use_paging 0 Paging for flat tables used (1) or not (0)

[Former Member]

Thanks for your response.Please check attached text file st22.txt

[mercedes_fernandez]

Hello,

KBA 2180736 - TSV_TNEW_PAGE_ALLOC_FAILED describes how to analyze this kind of dump
You could find specific platform information in the BC-CST wiki page

http://wiki.scn.sap.com/wiki/x/24kdGQ#content

kind regards,
Mercedes

[manuel_garcia]

Hi Godakrishna,

Checking the dump it seems more an application issue in SRM system, while using the Shopping Cart transactions ,than a problem to be solved modifiying sap memory parameters.

Program SAPLBBP_PD is consuming a lot of memory (around 10 GB) when the memory dump occurs. Normally the reason has to be analyzed from SRM point of view. You can also refer to the following documentation:

User can't access a POWL Query

Kind regards,

Manuel

[raquel_gomez]

Hi Godakirshna,

TSV_TNEW_PAGE_ALLOC_FAILED dump normally appears when a work process has taken its quota of Extended Memory and its quota of Heap Memory and has no more memory to allocate.
This is detailed on #scenario1 from KBA "2180736 - TSV_TNEW_PAGE_ALLOC_FAILED ".

Checking the dump, WP took 2GB of EM and 8GB of HEAP.
The amount of storage space (in bytes) filled at termination time was:
Extended memory (EM)........... 2002738624
Assigned memory (HEAP)......... 8000009088

This can be fixed either increasing memory resources, so that more memory can be allocated; or reducing selection criteria so that less memory is requested.

Regards,
Raquel

[Matt_Fraser]

Hi Godakrishna,

In addition to what Mercedes posted (excellent resources), another Note you should make yourself familiar with:

Zero Administration Memory Management for the ABAP Server

Also, as the memory dump is occurring in an SRM system, while using the Shopping Cart transactions, check out:

Poor Performance When Opening Shopping Cart

POWL Partial Refresh: Memory Dump TSV_TNEW_PAGE_ALLOC_FAILED

Are you using the "New UI" for shopping carts? There are lots of memory Notes that need to be applied in that circumstance, as I recall.

One of the bottom lines, however, is that mostly you shouldn't set too many of those memory parameters, at least initially, and instead let them default based upon a correct setting of PHYS_MEMSIZE and having a large enough page file in Windows.

As an example, this is how our production SRM system is configured (it's running SRM 7.0 EhP3, with the New UI, and therefore on NetWeaver 7.40, on MSSQL 2012 and Windows 2012 R2). We have 64 GB of RAM on the server, with database and application server installed on a single host. We configured the pagefile to be 40 GB in size. We fixed the SQL Server memory at 26,200 MB. In the instance profile, we set:

PHYS_MEMSIZE = 33157
gw/max_conn = 2000
gw/max_overflow_size = 25000000
rdisp/max_comm_entries = 2000
rdisp/task_limit = 2000
rdisp/tm_max_no = 2000
rsdb/cua/buffersize = 5000
zcsa/presentation_buffer_area = 20000000

That's it. This system runs without any swaps in any of the buffers, and we aren't getting memory overflow dumps. Performance is acceptable, though we had to apply a whole bunch of Notes to get the New UI to work well.

Cheers,
Matt

p.s. I adjusted the tags on your question to reflect that it's an SRM system.

[antoine_foucault]

Hi Godakirshna,

As mentioned by Ashutosh, could post the dump in ST22?

Most likely it is due to an infinite loop somewhere in your report.

Antoine

[former_member185239]

Hi,

Can you please paste the st22 dump.

With Regards
Ashutosh Chaturvedi