SAP starts too slow after upgrade

[Former Member]

Hi,

I just finished upgrading our system from 4.6D to ECC 6.But when I try to start the sap, I noticed that the work process were starting very slow (more than 1 hour to start),Here the log file that I get from dev_w log :

M Fri Feb 1 22:30:08 2008

M SHM_PRES_BUF (addr: 700000300000000, size: 1677312000)

M SHM_ROLL_AREA (addr: 700000370000000, size: 268435456)

M

M Fri Feb 1 22:30:09 2008

M SHM_PAGING_AREA (addr: 700000380000000, size: 268435456)

M SHM_ROLL_ADM (addr: 700000390000000, size: 2672386)

M SHM_PAGING_ADM (addr: 7000003a0000000, size: 656416)

M ThCreateNoBuffer allocated 324144 bytes for 1000 entries at 7000003b0000000

M ThCreateNoBuffer index size: 3000 elems

M ThCreateVBAdm allocated 7440 bytes (50 server) at 7000003c0000000

X EmInit: MmSetImplementation( 2 ).

X MM global diagnostic options set: 0

X

X Fri Feb 1 22:30:10 2008

X <ES> client 8 initializing ....

X Using implementation std

X ES initialized.

B db_con_shm_ini: WP_ID = 8, WP_CNT = 32, CON_ID = -1

B dbtbxbuf: Buffer TABL (addr: 700000430000100, size: 64000000, end: 700000433d09100)

B

B Fri Feb 1 22:30:11 2008

B dbtbxbuf: Buffer TABLP (addr: 700000440000100, size: 16776192, end: 700000440fffd00)

B dbexpbuf: Buffer EIBUF (addr: 700000460000108, size: 30720000, end: 700000461d4c108)

B dbexpbuf: Buffer ESM (addr: 700000470000108, size: 4194304, end: 700000470400108)

B dbexpbuf: Buffer CUA (addr: 700000480000108, size: 16776192, end: 700000480fffd08)

B

B Fri Feb 1 22:30:12 2008

B dbexpbuf: Buffer OTR (addr: 700000490000108, size: 4194304, end: 700000490400108)

M

M Fri Feb 1 22:30:18 2008

M CCMS: AlInitGlobals : alert/use_sema_lock = TRUE.

S

S Fri Feb 1 22:30:20 2008

S *** init spool environment

S initialize debug system

T Stack direction is downwards.

T debug control: prepare exclude for printer trace

T new memory block 18294edf0

S spool kernel/ddic check: Ok

S using table TSP02FX for frontend printing

S 4 spool work process(es) found

S frontend print via spool service enabled

S printer list size is 150

S printer type list size is 50

S queue size (profile) = 300

S hostspool list size = 3000

S option list size is 30

S found processing queue enabled

S found spool memory service RSPO-RCLOCKS at 7000004d0000060

S doing lock recovery

S setting server cache root

S found spool memory service RSPO-SERVERCACHE at 7000004d00008a8

S using messages for server info

S size of spec char cache entry: 165024 bytes (timeout 100 sec)

S size of open spool request entry: 1184 bytes

S immediate print option for implicitely closed spool requests is disabled

Any Idea how to solve this?

[Former Member]

Hi hazrienne,

DSPSYSVAL QBASACTLVL is interesting as well ...

Regards

Volker Gueldenpfennig, consolut international ag

http://www.consolut.de - http://www.4soi.de - http://www.easymarketplace.de

[Former Member]

Hi hazrienne,

first of all: Yes, starting of SAP is FAR slower with this kernel, than before ...

BUT:

- How many WPs are you running ?

- How many CPUs do you have ?

- What kind of CPUs ? 820, i5, i5+, i6 ?

- How much RAM do you have in QMACHPOOL and on the base pool ?

When did you really start SAP ? at 22:30 ? Or 1 hour earlier ?

... because the traces show only 1 minute ;-(

Regards

Volker Gueldenpfennig, consolut international ag

http://www.consolut.de - http://www.4soi.de - http://www.easymarketplace.de

[Former Member]

Hi,

The answers of your questions :

How many WPs are you running ?

31

How many CPUs do you have ?

1.5 (LPAR)

What kind of CPUs ? 820, i5, i5+, i6 ?

i550

How much RAM do you have in QMACHPOOL and on the base pool ?

16GB

I already applied the latest kernel patches (143) and the latest PTF patches, but returned the same results.

It starts after 1 hour later (cant post the whole log - too long).

Regards

[Former Member]

Hi hazrienne,

16 WPs are more than sufficient for 1.5 CPUs, but this should not be the issue ...

Regards

It starts after 1 hour later (cant post the whole log - too long).

=> it is not interesting, to see everything, but it would be interesting the first time stamp and the last one ...

So, again: When did you start ?????????

When you start - what happens then ? Is it 1h on 100% CPU ?

Or do we have 0% for 45 or 55 minutes and then 100% and then it really starts up ?

Volker Gueldenpfennig, consolut international ag

http://www.consolut.de - http://www.4soi.de - http://www.easymarketplace.de

[Former Member]

Hi,

the CPU were just 16 - 20 % only, It took 60 minutes just to make the wp status to SEMW