Enqueue Standalone vs Enqueue Work Process

[Former Member]

Dear Gurus,

I'm currently have this landscape:

Node 1

-

---

CI

DB

ASCS

Node 2

-

---

Dialog Instance

System is configured that when Node 1 down, CI, DB, and ASCS move to Node 2, and run from there.

The question is, do you need to set the Enqueue Work Process when you have run ASCS, or just put 0 in Enqueue WP in RZ10?

Thanks for help.

Regards,

[arun_lobo · Accepted solution]

Just looking at the profiles

ASCS profile

###################################

SAPSYSTEMNAME = SM1

SAPSYSTEM = 01

INSTANCE_NAME = ASCS01

DIR_CT_RUN = $(DIR_EXE_ROOT)/run

DIR_EXECUTABLE = $(DIR_INSTANCE)/exe

SAPLOCALHOST = up-ssm-cs

DIR_PROFILE = $(DIR_INSTALL)/profile

PF = $(DIRPROFILE)/SM1_ASCS01_up-ssm-cs

SETENV_00 = LD_LIBRARY_PATH=$(DIR_LIBRARY):%(LD_LIBRARY_PATH)

SETENV_01 = SHLIB_PATH=$(DIR_LIBRARY):%(SHLIB_PATH)

SETENV_02 = LIBPATH=$(DIR_LIBRARY):%(LIBPATH)

#----

---

1. Copy SAP Executables

#----

---

CPARG0 = list:$(DIRCT_RUN)/scs.lst

Execute_00 = immediate $(DIR_CT_RUN)/sapcpe$(FT_EXE) pf=$(_PF) $(_CPARG0)

#----

---

1. Start SAP message server

#----

---

MS = ms.sap$(SAPSYSTEMNAME)$(INSTANCE_NAME)

Execute_01 = local rm -f $(_MS)

Execute_02 = local ln -s -f $(DIR_EXECUTABLE)/msg_server$(FT_EXE) $(_MS)

Start_Program_00 = local $(_MS) pf=$(_PF)

#----

---

1. Start SAP enqueue server

#----

---

EN = en.sap$(SAPSYSTEMNAME)$(INSTANCE_NAME)

Execute_03 = local rm -f $(_EN)

Execute_04 = local ln -s -f $(DIR_EXECUTABLE)/enserver$(FT_EXE) $(_EN)

Start_Program_01 = local $(_EN) pf=$(_PF)

#######################################

enque replication profile

#######################################

sm1adm> less START_ERS11_up-ssm-enrep

SAPSYSTEMNAME = SM1

SAPSYSTEM = 11

INSTANCE_NAME = ERS11

SCSID = 01

DIR_CT_RUN = /usr/sap/SM1/SYS/exe/run

DIR_EXECUTABLE = $(DIR_INSTANCE)/exe

DIR_PROFILE = /sapmnt/SM1/profile

PF = $(DIRPROFILE)/SM1_ERS11_up-ssm-enrep

SETENV_00 = LD_LIBRARY_PATH=$(DIR_LIBRARY):%(LD_LIBRARY_PATH)

SETENV_01 = SHLIB_PATH=$(DIR_LIBRARY):%(SHLIB_PATH)

SETENV_02 = LIBPATH=$(DIR_LIBRARY):%(LIBPATH)

#----

---

1. Copy SAP Executables

#----

---

CPARG0 = list:$(DIREXECUTABLE)/ers.lst

Execute_00 = immediate $(DIR_EXECUTABLE)/sapcpe$(FT_EXE) $(_CPARG0) pf=$(_PF)

#----

---

1. start enqueue replication server

#----

---

ER = er.sap$(SAPSYSTEMNAME)$(INSTANCE_NAME)

Execute_01 = immediate rm -f $(_ER)

Execute_02 = local ln -s -f $(DIR_EXECUTABLE)/enrepserver $(_ER)

Restart_Program_00 = local $(_ER) pf=$(_PF) NR=$(SCSID)

~

now in the default profile we have

###################################

enque/serverhost up-ssm-cs

enque/serverinst 01

if you read above, up-ssm-cs in this case is the host that will be always alive and pointing to current central instance.

hope that helps

[Former Member · Accepted solution]

When you say current central instance, do you mean the ASCS instance or the DVEBMGS instance?

Regards,

[former_member188883 · Accepted solution]

Hi Lori,

Central instance here means DVEBMGS*

Regards,

Deepak Kori

[sunny_pahuja2 · Accepted solution]

Hi,

ASCS instance will be ASCS<nn>. CI is DVEBMGS<nn>, DI is D<nn>. when node 1 will be down, all users logon CI instance will be disconnected.

Thanks

Sunny

[Former Member · Accepted solution]

Hi,

You said before that when node 1 is down (hardware down), node 2 will work.

But if the DB is in node 1, the DB will then become unavailable, until it is started on node 2.

How come node 2 will still work without DB? Is there a parameter to set the maximum time SAP instance should wait for database, before disconnecting the user logged on in node 2?

Regards,

[sunny_pahuja2 · Accepted solution]

Hi,

You are right. There will be small disconnect fron DI as well till the time DB and ASCS instance will not be started on node 2. But there are certain parameters which you can define to increase reconnect time. I have not tried this in Unix cluster but this works in MSCS cluster.

Check SAP Note 24806 - Database Reconnect technical details and settings.

Thanks

Sunny

[sunny_pahuja2]

Hi,

In case of HA systems, Enqueue server lies in ASCS. So, if node 1 down then ASCS moves to node 2 and it will provide Enqueue service there. So, you need to set it in instance profile in RZ10.

Thanks

Sunny

[Former Member]

Hi,

This one confuses me for long.

If node 1 down, which is where the DB is also located, doesn't it mean that the whole system will go down (ERS become useless to) ?

So what's the use of ASCS instead of Classic Enqueue WP?

Best Regards,

[sunny_pahuja2]

Hi,

If your node 1 is down then ASCS and DB will move to node 2. Since ASCS will provide Enqueue services. Only CI will die as node 1 is down. Then DI on node 2 will work.

Thanks

Sunny

[Former Member]

When DB switch to node 2, won't be there short time system down?

This also means all user context currently logged in CI will be lost, user will be kicked out of the system.

Dialog in node 2 can't find the DB until it is started on node 2, hence the user currently logged in to node 2 will be out of the system also.

Regards,