even after restart no logon possible: Max. number internal sessions reached

[rdiger_hckel]

When trying to logon via SAPGUI one gets only the message:

Maximum number of internal sessions reached

(in our case the message is in German: Maximale Anzahl interner Modi erreicht)

As this is a new system that has just been installed here is the short

system history so far:

• installation of SRM 7.0 EHP1 on IBM i V6R1

• deployment of latest kernel patches: SAPEXE/SAPEXEDB #120 (7.20 kernel64bit unicode), ILE #120

• STMS configuration and some minor setup work in the new system (SCC4,

SPAD, SM65 and so on)

• latest SPAM/SAINT update: 7.02/0045

• according to SAP note 819722 the support package level of the

installation DVDs is rather old so I created a rather long queue of to be

imported ABAP support packages (PI_BASIS 702 #4-10, SAP_ABA 702 #4-10,

SAP_AP 700 #19-16, SAP_BASIS 702 #4-10, SAP_BS_FND 702 #4-10, SAP_BW 702#4-10, SRM_EXPL 701 #1-4, SRM_EXT 701 #1-5, SRM_PLUS 701 #1-5, SRM_SERVER 701 #1-5)

the SPAM import ran throughout the whole Friday and must have been stopped/hanging since late Friday afternoon

in phase TBATG CREATION OF ENQUEUE

I have restarted the system but the logon error continues.

Since Friday 17:41 o'clock every 5 minutes the following message appearsin the SLOG file:

WARNING: System SR1. Warning. 20120127174114 :

WARNING: Background job RDDIMPDP could not be started or terminated

abnormally.

Please check that the R/3 system is running.

Please check the system. Use transactions SM21, SM37, SM50.

WARNING: (This warning is harmless if no further warnings follow.)

The SPAM window was "frozen" so on Saturday morning I restarted the SAP

system.

content of table TRBAT:

select * from trbat

TRKORR FUNCTION RETCODE TIMESTMP CLIENT LANGU VARIANT ARGUMENT LOGNAME DATNAME

HEADER Y B 20120127173706 :D:T:S:tmp:F:N120127.SR1

1 RECORD(S) SELECTED.

This morning I deployed the very latest kernel patches: ILE #121, dw #121

and lib_dbsl #121 and even deleted the SQL packages (because of the new

lib_dbsl version) -> no success

command:

./tp checkimpdp SR1 pf=/usr/sap/trans/bin/TP_DOMAIN_SRM.PFL

delivers:

System SR1: Background job RDDIMPDP is scheduled event periodic.

select * from tbtco where jobname = 'RDDIMPDP' and status = 'S'

RDDIMPDP 10392900 %NEWSTEP

1

20120127 103929 DDIC

20100518 094936 DDIC 20120127 103929 SAPSYS

00 00 000 00

00 X S

C 000 0 0

0 0

SAP_TRIGGER_RDDIMPDP

C 0 09493601

0 0

1 RECORD(S) SELECTED.

The problem is that somehow some internal modes are occupied so other processes (for instance, RDDIMPDP batch job) canot be called.

M Tue Jan 31 10:21:21 2012

M ***LOG R0A=> ThCreateIntMode, max imodes (09) [thxxmode.c 396]

-> this entry is written permanently in three dev_wnn trace files

[Former Member]

Hi Rudiger,

perhaps posting in this forum will be more helpful?

[Former Member]

Hello

Logon to the DB

sqlplus "/as sysdba"

and increase the parameter values of processes and sessions Then restart both SAP and oracle.

Thanks & regards

bala

[rdiger_hckel]

Hello bala,

thanks for your quick response.

This system is running on iSeries ("on IBM i V6R1") so we don't have an Oracle database.

What I would like to know is if there is some database table in which these (or the number of) "internal (SAP) modes" are stored...

Kind regards

Rüdiger

[Former Member]

Hello

Please check this SAP note

Note 1035714 - Many TEXTENV_UNICODE_LANGU_INVALID short dumps

Thanks & regards

bala

[rdiger_hckel]

Yes, I also had found this SAP note but as I'm using a 720 kernel (unicode 64bit) it does not fit.

External CPIC clients? Well, as I said this is afreshly installed system, currently only the SPAM transaction tries to do some stuff with the system.

And

M Tue Jan 31 13:30:07 2012

M ***LOG R0A=> ThCreateIntMode, max imodes (09) [thxxmode.c 396]

is the only entry that's been added to the trace file.

I also stopped the system, cleared the shared memory and started again - but no success.

[Former Member]

Hello

Try renaming the latest file in /usr/sap/trans/tmp.

Thanks & regards

bala

[rdiger_hckel]

I changed the name of the newest file from

SAPRUPGM.SR1

to

SAPRUPGM.SR1.save

and restarted the system - still no success.

Even changed the file names of UMODPROT.SR1, IMODPROT.SR1 and AMODPROT.SR1 in the same directory but no success.

[rdiger_hckel]

As I'm running out of time I decided to delete the existing installation and start a new installation from scratch.

[Former Member]

Hi Rüdiger,

So you never found out the root cause of this?

I ran into same problem with a CRM 7.01 / IBM i installation.

Did you do something differently?

Thanks.

Samuli

[Cruizs]

Hello,

   I have the same error. I'm updating an ERP 6.0  EHP5 SPS3 to SPS8. The system is running on iSeries.  Has anyone found a solution?.

[rdiger_hckel]

Hello,

when installing the production system I had the same error.

"As I'm running out of time I decided to delete the existing installation and start a new installation from scratch."

After the installation I added a few support packages to the queue but I still had that error.

So after a new installation run I think the only thing I forgot so far was to execute SE06 (Perform Post-Installation Actions) after the installation. After that it worked fine.

Maybe some configuration in the transport management system was missing, I only did the configuration via STMS.

But to be honest I really don't know the cause of this problem.

Hope this helps,

   Rüdiger

[Former Member]

Hi,

Just for the record: our problem was an interrupted basis support stack.

After finishing the stack import, the error disappeared.

I was able to log on to the system by entering "/SUPPORTBIT_ON=EMERGENCY_REPAIR" in my SAPGUI startup string, according to SAP note 396802.

After successfully logging in to the system, I was able to complete the stack.

BR,

Samuli

[Former Member]

That's it.

Just set "/SUPPORTBIT_ON=EMERGENCY_REPAIR" in SAP GUI, logon to the System and in SE38 execute Report: SDB4GEN

than you can continue the SP update.

Works only for system i!