Dispatcher Stopped

[former_member182034]

Dear All,

I have installed DEV and QAS on same system(Window server 2003) with oracle 10.2 and ECC6.0 SR3.

DEV server is running fine while i try to run QAS after 1 month then Dispatcher went to stopped while database of QAS is successfully startup and shutdown.

Please have a look on DEV_disp logs and guide me

-

---

trc file: "dev_disp", trc level: 1, release: "700"

-

---

sysno 01

sid QAS

systemid 562 (PC with Windows NT)

relno 7000

patchlevel 0

patchno 144

intno 20050900

make: multithreaded, Unicode, 64 bit, optimized

pid 3632

Thu Jun 11 17:53:49 2009

kernel runs with dp version 232000(ext=109000) (@(#) DPLIB-INT-VERSION-232000-UC)

length of sys_adm_ext is 576 bytes

• SWITCH TRC-HIDE on ***

***LOG Q00=> DpSapEnvInit, DPStart (01 3632) [dpxxdisp.c 1243]

shared lib "dw_xml.dll" version 144 successfully loaded

shared lib "dw_xtc.dll" version 144 successfully loaded

shared lib "dw_stl.dll" version 144 successfully loaded

shared lib "dw_gui.dll" version 144 successfully loaded

shared lib "dw_mdm.dll" version 144 successfully loaded

rdisp/softcancel_sequence : -> 0,5,-1

use internal message server connection to port 3901

Thu Jun 11 17:53:55 2009

• WARNING => DpNetCheck: NiAddrToHost(1.0.0.0) took 5 seconds

***LOG GZZ=> 1 possible network problems detected - check tracefile and adjust the DNS settings [dpxxtool2.c 5371]

MtxInit: 30000 0 0

DpSysAdmExtInit: ABAP is active

DpSysAdmExtInit: VMC (JAVA VM in WP) is not active

DpIPCInit2: start server >DEVQAS_QAS_01 <

DpShMCreate: sizeof(wp_adm) 25168 (1480)

DpShMCreate: sizeof(tm_adm) 5652128 (28120)

DpShMCreate: sizeof(wp_ca_adm) 24000 (80)

DpShMCreate: sizeof(appc_ca_adm) 8000 (80)

DpCommTableSize: max/headSize/ftSize/tableSize=500/16/552064/552080

DpShMCreate: sizeof(comm_adm) 552080 (1088)

DpSlockTableSize: max/headSize/ftSize/fiSize/tableSize=0/0/0/0/0

DpShMCreate: sizeof(slock_adm) 0 (104)

DpFileTableSize: max/headSize/ftSize/tableSize=0/0/0/0

DpShMCreate: sizeof(file_adm) 0 (72)

DpShMCreate: sizeof(vmc_adm) 0 (1864)

DpShMCreate: sizeof(wall_adm) (41664/36752/64/192)

DpShMCreate: sizeof(gw_adm) 48

DpShMCreate: SHM_DP_ADM_KEY (addr: 0000000016E70050, size: 6348592)

DpShMCreate: allocated sys_adm at 0000000016E70050

DpShMCreate: allocated wp_adm at 0000000016E72150

DpShMCreate: allocated tm_adm_list at 0000000016E783A0

DpShMCreate: allocated tm_adm at 0000000016E78400

DpShMCreate: allocated wp_ca_adm at 00000000173DC2A0

DpShMCreate: allocated appc_ca_adm at 00000000173E2060

DpShMCreate: allocated comm_adm at 00000000173E3FA0

DpShMCreate: system runs without slock table

DpShMCreate: system runs without file table

DpShMCreate: allocated vmc_adm_list at 000000001746AC30

DpShMCreate: allocated gw_adm at 000000001746ACB0

DpShMCreate: system runs without vmc_adm

DpShMCreate: allocated ca_info at 000000001746ACE0

DpShMCreate: allocated wall_adm at 000000001746ACF0

MBUF state OFF

DpCommInitTable: init table for 500 entries

ThTaskStatus: rdisp/reset_online_during_debug 0

EmInit: MmSetImplementation( 2 ).

MM global diagnostic options set: 0

<ES> client 0 initializing ....

<ES> InitFreeList

<ES> block size is 4096 kByte.

<ES> Info: em/initial_size_MB( 24574MB) not multiple of em/blocksize_KB( 4096KB)

<ES> Info: em/initial_size_MB rounded up to 24576MB

Using implementation view

<EsNT> Using memory model view.

<EsNT> Memory Reset disabled as NT default

<ES> 6143 blocks reserved for free list.

ES initialized.

***LOG Q0I=> NiIBindSocket: bind (10013: WSAEACCES: Permission denied) [nixxi.cpp 3227]

• ERROR => NiIBindSocket: SiBind failed for hdl 0 / sock 140

(SI_EPORT_INUSE/10013; I4; DG; 127.0.0.1:3201) [nixxi.cpp 3227]

• ERROR => DpCommInit: NiDgHdlBindName failed: -4 [dpxxdisp.c 9483]

• DP_FATAL_ERROR => DpSapEnvInit: DpCommInit

• DISPATCHER EMERGENCY SHUTDOWN ***

increase tracelevel of WPs

NiWait: sleep (10000ms) ...

NiISelect: timeout 10000ms

NiISelect: maximum fd=141

NiISelect: read-mask is NULL

NiISelect: write-mask is NULL

Thu Jun 11 17:54:05 2009

NiISelect: TIMEOUT occured (10000ms)

dump system status

Workprocess Table (long) Thu Jun 11 12:54:05 2009

========================

No Ty. Pid Status Cause Start Err Sem CPU Time Program Cl User Action Table

-

---

0 DIA -1 Run no 0 0 0

1 DIA -1 Run no 0 0 0

2 DIA -1 Run no 0 0 0

3 DIA -1 Run no 0 0 0

4 DIA -1 Run no 0 0 0

5 DIA -1 Run no 0 0 0

6 DIA -1 Run no 0 0 0

7 DIA -1 Run no 0 0 0

8 DIA -1 Run no 0 0 0

9 DIA -1 Run no 0 0 0

10 UPD -1 Run no 0 0 0

11 ENQ -1 Run no 0 0 0

12 BTC -1 Run no 0 0 0

13 BTC -1 Run no 0 0 0

14 BTC -1 Run no 0 0 0

15 SPO -1 Run no 0 0 0

16 UP2 -1 Run no 0 0 0

Dispatcher Queue Statistics Thu Jun 11 12:54:05 2009

===========================

--------

---

+

---

+

---

+

---

+--

---

+

Typ

now

high

max

writes

reads

--------

---

+

---

+

---

+

---

+--

---

+

NOWP

0

0

2000

0

0

--------

---

+

---

+

---

+

---

+--

---

+

DIA

0

0

2000

0

0

--------

---

+

---

+

---

+

---

+--

---

+

UPD

0

0

2000

0

0

--------

---

+

---

+

---

+

---

+--

---

+

ENQ

0

0

2000

0

0

--------

---

+

---

+

---

+

---

+--

---

+

BTC

0

0

2000

0

0

--------

---

+

---

+

---

+

---

+--

---

+

SPO

0

0

2000

0

0

--------

---

+

---

+

---

+

---

+--

---

+

UP2

0

0

2000

0

0

--------

---

+

---

+

---

+

---

+--

---

+

max_rq_id 0

wake_evt_udp_now 0

wake events total 0, udp 0 ( 0%), shm 0 ( 0%)

since last update total 0, udp 0 ( 0%), shm 0 ( 0%)

Dump of tm_adm structure: Thu Jun 11 12:54:05 2009

=========================

Term uid man user term lastop mod wp ta a/i (modes)

Workprocess Comm. Area Blocks Thu Jun 11 12:54:05 2009

=============================

Slots: 300, Used: 0, Max: 0

--------

---

+

---

+--

---

+

id

owner

pid

eyecatcher

--------

---

+

---

+--

---

+

NiWait: sleep (5000ms) ...

NiISelect: timeout 5000ms

NiISelect: maximum fd=141

NiISelect: read-mask is NULL

NiISelect: write-mask is NULL

Thu Jun 11 17:54:10 2009

NiISelect: TIMEOUT occured (5000ms)

DpHalt: shutdown server >DEVQAS_QAS_01 < (normal)

DpJ2eeDisableRestart

Switch off Shared memory profiling

ShmProtect( 57, 3 )

ShmProtect(SHM_PROFILE, SHM_PROT_RW

ShmProtect( 57, 1 )

ShmProtect(SHM_PROFILE, SHM_PROT_RD

DpWakeUpWps: wake up all wp's

[DpWpDied] W0 handle not set (PID:-1)

[DpWpDied] W1 handle not set (PID:-1)

[DpWpDied] W2 handle not set (PID:-1)

[DpWpDied] W3 handle not set (PID:-1)

[DpWpDied] W4 handle not set (PID:-1)

[DpWpDied] W5 handle not set (PID:-1)

[DpWpDied] W6 handle not set (PID:-1)

[DpWpDied] W7 handle not set (PID:-1)

[DpWpDied] W8 handle not set (PID:-1)

[DpWpDied] W9 handle not set (PID:-1)

[DpWpDied] W10 handle not set (PID:-1)

[DpWpDied] W11 handle not set (PID:-1)

[DpWpDied] W12 handle not set (PID:-1)

[DpWpDied] W13 handle not set (PID:-1)

[DpWpDied] W14 handle not set (PID:-1)

[DpWpDied] W15 handle not set (PID:-1)

[DpWpDied] W16 handle not set (PID:-1)

Stop work processes

Terminate gui connections

wait for end of work processes

not attached to the message server

cleanup EM

EsCleanup ....

EmCleanup() -> 0

Es2Cleanup: Cleanup ES2

***LOG Q05=> DpHalt, DPStop ( 3632) [dpxxdisp.c 10421]

Good Bye .....

i saw the DISPATCHER EMERGENCY SHUTDOWN in dev_disp log while i checked the following entries

which are exist in the service file.

sapdp00 3200/tcp

sapdp01 3201/tcp

sapmsDEV 3600/tcp

sapmsQAS 3601/tcp

Regards,

[markus_doehr2 · Accepted solution]

> ***LOG Q0I=> NiIBindSocket: bind (10013: WSAEACCES: Permission denied) [nixxi.cpp 3227]

> *** ERROR => NiIBindSocket: SiBind failed for hdl 0 / sock 140

> (SI_EPORT_INUSE/10013; I4; DG; 127.0.0.1:3201) [nixxi.cpp 3227]

> *** ERROR => DpCommInit: NiDgHdlBindName failed: -4 [dpxxdisp.c 9483]

> *** DP_FATAL_ERROR => DpSapEnvInit: DpCommInit

> *** DISPATCHER EMERGENCY SHUTDOWN ***

What happened prior to this? Did you install updates/patches for the OS or any other software installed?

Markus

[former_member182034 · Accepted solution]

thanks Markus for replies,

QAS was working fine before one month and was no need of QAS server because here BBP phase is running, so i stopped and shutdown the QAS server. i did not perform any patch or updates, i simply installed the ISA firewall client on this Server and i disabled the ISA firewall client and try to start the server but dispatcher stopped agian.

Regards,

[markus_doehr2 · Accepted solution]

The error message points exactly in that direction: The system is not allowed to connect to the dispatcher port. I'm not sure how ISA handles this but I remember, that a deactivated firewall on Windows sometimes leads to blocking all connections.

Markus

[former_member182034 · Accepted solution]

thanks Markus

one more thing.

sorry here i am putting one more issue.. i can say other question but it related to it.

i was facing some red swapping on DEV server,, so increased and put some parameter in instance profiles those are following:

1. update V1 work process = 3 from 1
2. Update V2 work process = 2 from 1 
2. abap/buffersize = 512000   from 300000
3. PHYS_MEMSIZE = 7168   from 24574   
4. zcsa/presentation_buffer_area = 6400000 from 4400000

DEV and QAS on same machine with 24GB.  i did give 10GB to dev and rest for QAS
while 20 Users are working on DEV

may be this cause of this issue?

Regards,

[markus_doehr2 · Accepted solution]

> i was facing some red swapping on DEV server,, so increased and put some parameter in instance profiles those are following:

> may be this cause of this issue?

No.

The error message clearly indicates that it's a problem binding the network socket:

> ***LOG Q0I=> NiIBindSocket: bind (10013: WSAEACCES: Permission denied) http://nixxi.cpp 3227

> *** ERROR => NiIBindSocket: SiBind failed for hdl 0 / sock 140

> SI_EPORT_INUSE/10013; I4; DG; 127.0.0.1:3201) http://nixxi.cpp 3227

I suggest the following:

Download TCPview from http://technet.microsoft.com/en-us/sysinternals/bb897437.aspx and see which program occupies port 3201.

Markus

[former_member182034 · Accepted solution]

Hi markus,

there are status of port 3201 in TCPView is:

Process: disp+work.exe:4056

Protocol: UDP

Local Address: DEVQAS:3201 (DEVQAS is host name)

Remote address: .

State: ...............

while in Service file:

sapmsDEV	3600/tcp	# SAP System Message Server Port
sapmsQAS	3601/tcp	# SAP System Message Server Port
sapdp00  3200/tcp
sapdp01  3201/tcp
sapdp02  3202/tcp
sapdp03  3203/tcp
sapdp04  3204/tcp

in TCPview 3201 is UDP port while in service file this is TCP

Regards,

[markus_doehr2 · Accepted solution]

did you always stop the system before trying to restart it?

I suggest you stop it, check again with TCPview if there are open ports and then try to start it again. If that fails I would reboot the machine and before starting the instance again use again the tool to see if the port is free.

Markus

[former_member182034 · Accepted solution]

Thanks markus,

Problem solved with restart the system

may i shutdown both(DEV & QAS) database from DEVadm user name or i have to shutdown QAS db with qasadm and DEV db with devadm.....?????

here require ur prodigy look on my other post Swapping increase.

I m really worry about Red Swapping plzzzzzzzzzzzzz have a look and i already mention regarding Swap in this post.

Regards,

Edited by: majamil on Jun 11, 2009 8:24 PM

[markus_doehr2 · Accepted solution]

> Problem solved with restart the system.

Ok - good to hear it's working again - although that solution (the reboot) for me would be not satisfying. It should not be necessary to reboot an OS to get applications working again but that's just my EUR 0.02.

> here require ur prodigy look on my other post Swapping increase.

> I m really worry about Red Swapping plzzzzzzzzzzzzz have a look and i already mention regarding Swap in this post.

The "swapping" in ST02 means, that e. g. the program was not found in the buffer and must be loaded from the database.

How much you actually need depends heavily on what is done with the system. If you have many different programs running (so different 'modules' if you want) 512 MB is still not enough.

In our production environment we use 3,000,000 for abap/buffersize because we also import lots of transports that import new programs and invalidate the buffers and fragment them.

The number of work processes doesn't count much in the buffer area, they all connected to the shared memory and use the SAME buffers.

Markus

[Former Member]

possible network problems detected - check tracefile and adjust the DNS settings

Check your network connectivity, firewall rules