on 06-30-2008 2:04 PM
Dear all,
System details:
Solaris 10 with 2 zones, oracle 10g, ECC 6 and NW 7.
I have 2 zones. In my first zone, i installed 2 sap instances, one instance is ECC 6 (abap stack only) and another with NW 7 (as abap for bw). Both intances have their own databases and oracle instances.
When i start Oracle instances and databases no problems. But when i start sap, i can start only one instance. if i start 2nd intance it give me the error.
dev_w0
trc file: "dev_w0", trc level: 1, release: "700"
-
*
ACTIVE TRACE LEVEL 1
ACTIVE TRACE COMPONENTS all, MJ
*
M sysno 21
M sid BIQ
M systemid 372 (Solaris on x86_64 CPU)
M relno 7000
M patchlevel 0
M patchno 144
M intno 20050900
M make: single threaded, Unicode, 64 bit, optimized
M pid 4417
M
M
M Mon Jun 30 15:09:28 2008
M kernel runs with dp version 232000(ext=109000) (@(#) DPLIB-INT-VERSION-232000-UC)
M length of sys_adm_ext is 576 bytes
M ***LOG Q01=> ThInit, WPStart (Workproc 0 1 4417) [thxxhead.c 1281]
M ThInit: running on host sapqasabap
I *** ERROR => e=28 semget(22128,1,2047) (28: No space left on device) [semux.c 928]
M *** ERROR => ThrRegisterSem: SemInit SEM_SAPCOM_2 [thxxrun1.c 626]
M *** ERROR => ThCallHooks: event handler ThrRegisterSem for event BEFORE_DB_CONNECT failed [thxxtool3.c 261]
M in_ThErrHandle: 1
M *** ERROR => ThCallHooks: hook failed (step 1, th_errno 5, action 3, level 1) [thxxhead.c 10468]
M
M Info for wp 0
M
M stat = WP_RUN
M waiting_for = NO_WAITING
M reqtype = DP_RQ_DIAWP
M act_reqtype = NO_REQTYPE
M rq_info = 0
M tid = -1
M mode = 255
M len = -1
M rq_id = 65535
M rq_source =
M last_tid = 0
M last_mode = 0
M semaphore = 0
M act_cs_count = 0
M csTrack = 0
M csTrackRwExcl = 0
M csTrackRwShrd = 0
M mode_cleaned_counter = 0
M control_flag = 0
M int_checked_resource(RFC) = 0
M ext_checked_resource(RFC) = 0
M int_checked_resource(HTTP) = 0
M ext_checked_resource(HTTP) = 0
M report = > <
M action = 0
M tab_name = > <
M attachedVm = no VM
M PfStatDisconnect: disconnect statistics
M Entering TH_CALLHOOKS
M ThCallHooks: call hook >ThrSaveSPAFields< for event BEFORE_DUMP
M *** ERROR => ThrSaveSPAFields: no valid thr_wpadm [thxxrun1.c 724]
M *** ERROR => ThCallHooks: event handler ThrSaveSPAFields for event BEFORE_DUMP failed [thxxtool3.c 261]
M Entering ThSetStatError
M ThIErrHandle: do not call ThrCoreInfo (no_core_info=0, in_dynp_env=0)
M Entering ThReadDetachMode
M call ThrShutDown (1)...
M ***LOG Q02=> wp_halt, WPStop (Workproc 0 4417) [dpuxtool.c 260]
This error is same if i start any 2nd instance. The first instance (anyone) has no problem
Can anyone guide me.
Thanks
Did you create project files (in the zone) for the instances and assigned the users?
Markus
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
yes i created, the files are here:
/etc/project
system:0::::
user.root:1::::
noproject:2::::
default:3::::
group.staff:10::::
ECQ:202:SAP Sytem ECC:ecqadm,oraecq::process.max-sem-nsems=(priv,2048,deny);project.max-sem-ids=(priv,1024,deny);project.max-shm-ids=(priv,256,deny);project.max-shm-memory=(priv,18446744073709551615,deny)
BIQ:203:SAP System BIQ:biqadm,orabiq::process.max-sem-nsems=(priv,2048,deny);project.max-sem-ids=(priv,1024,deny);project.max-shm-ids=(priv,256,deny);project.max-shm-memory=(priv,18446744073709551615,deny)
/etc/user_attr
Copyright 2007 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
#
/etc/user_attr
#
execution attributes for profiles. see user_attr(4)
#
#ident "@(#)user_attr 1.1 07/01/31 SMI"
#
#
adm::::profiles=Log Management
lp::::profiles=Printer Management
postgres::::type=role;profiles=Postgres Administration,All
root::::auths=solaris.*,solaris.grant;profiles=Web Console Management,All;lock_after_retries=no
prdecq::::project=ECQ
oraecq::::project=ECQ
prdbiq::::project=BIQ
orabiq::::project=BIQ
Thanks
there was mistake in the file /etc/user_attr which i have corrected:
old:
prdecq::::project=ECQ
oraecq::::project=ECQ
prdbiq::::project=BIQ
orabiq::::project=BIQ
New:
ecqadm::::project=ECQ
oraecq::::project=ECQ
biqadm::::project=BIQ
orabiq::::project=BIQ
Then i saved, but the 2nd instance is still not coming up.
I think your shm-ids are too less.
The shm settings are upper limits, so it's no problem if you increase them more. I use the following settings:
projadd -p <proj-ID> -c "SAP System" -U ora<sid>,<sid>adm,root \
-K 'project.max-sem-ids=(priv,4096,deny)' \
-K 'process.max-sem-nsems=(priv,4096,deny)' \
-K 'process.max-file-descriptor=(priv,1024000,deny)' \
-K 'project.max-shm-ids=(priv,4096,deny)' \
-K 'project.max-shm-memory=(priv,18446744073709551615,deny)' \
SAP<SID>
Markus
If you decrease your Oracle processes to 60 and it works then, then it's an indication for a too low value of some of the parameters. 256 shm-ids are very low and may be used already from other programs. It's possible that you will get other errors during backup or other activities if you don't increase the value.
Markus
Can i use the values mentioned by you. I mean i have 10g ram and 48g swap.2 zones, zone 1 contain 2 abap installations and zone 2 contain 1 java installation (BI java, EP). How can i calculate to check my values.
Secondly to change these values i have to use gedit /etc/project or through the command. if command, can you tell me exactly.
There are 2 project in one zone and one project in another zone.
Thanks
hi dear,
I have made the same settings as mentioned as sample by you only for 2 abap systems in zone1 . After the succssful installations, when i restarted the systems the java is not coming up. and other 2 abap systems are running fine but they are very slow. I mean the whole OS is slow, even if i open a terminal it takes little time.
Should i remove shm parameters from /etc/system? or not?
or How can i check the performance, My ram is 10g and swap 48 g with 2 CPU dual core.
Thanks
i have adopted the values only and i am using the users of my own sid which i have maintained in /etc/user_attr.
ecdadm::::project=ECD
oraecd::::project=ECD
bidadm::::project=BID
orabid::::project=BID
/etc/project
ECD:201:SAP ECC System:ecdadm,oraecd::process.max-sem-nsems=(priv,4096,deny);project.max-sem-ids=(priv,4096,deny);project.max-shm-ids=(priv,4096,deny);project.max-shm-memory=(priv,18446744073709551615,deny)
BID:202:SAP Sytem BID:bidadm,orabid::process.max-sem-nsems=(priv,4096,deny);project.max-sem-ids=(priv,4096,deny);project.max-shm-ids=(priv,4096,deny);project.max-shm-memory=(priv,18446744073709551615,deny).
After that i just started sap system with database only. i didnt start the zone and i didnt start the OS. Should i or not?
Is my ram is sufficient to run 3 systems with 3 oracles instance (3 databases) 2 zones. swap is 48g
Thanks
looks all good!
> After that i just started sap system with database only. i didnt start the zone and i didnt start the OS. Should i or not?
No need to boot the zones.
> Is my ram is sufficient to run 3 systems with 3 oracles instance (3 databases) 2 zones. swap is 48g
It depends how much data you gave each Oracle instance.
Markus
I am using the default oracle parameters created during the installations. One system is ECC dev with 50 users, one with BW dev with 3-4 users and one with Portal dev (2-3) users. Using the default initsid.ora (no change)
during my old installationss, the oracle checker asked some parameter to be set. So for this reason in my current installation i used the same parameter in /etc/system in global
et rlim_fd_cur=8192
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10
set semsys:seminfo_semmns=1024
set semsys:seminfo_semmsl=256
set semsys:seminfo_semmni=100
set semsys:seminfo_semvmx=32767
set noexec_user_stack=1
Should i remove these parameters or not?
Thanks
> I am using the default oracle parameters created during the installations. One system is ECC dev with 50 users, one with BW dev with 3-4 users and one with Portal dev (2-3) users. Using the default initsid.ora (no change)
You configure the amount of memory - you are asked. Usually sapinst uses 50 % of the memory for the database instance. If you install three instances you will overcommit your memory.
> during my old installationss, the oracle checker asked some parameter to be set. So for this reason in my current installation i used the same parameter in /etc/system in global
Yes - that expected behaviour. You can ignore those warnings.
> Should i remove these parameters or not?
Yes - remove them but leave the rlim_cur_fd in place.
Markus
Markus,
I'm installing EHP4 with oracle 10.2.0.2 and Solaris 10.
When the EHPinst creates the shadow instance and tried to start it. I got the folowing error:
ERROR => e=28 semget(20038,1,2016) (28: No space left on device) [semux.c 1092]
ERROR => CCMS: SemInit failed [alxxappl.c 610]
ERROR => CCMS: sAlInit: could not create/attach shared memory key 13 (rtc = 255) [alxx.c 1187]
ERROR => AlDpAdmInit: sAlInit rtc = 255 [alxx.c 514]
ERROR => DpLoopInit: AlDpAdmInit [dpxxdisp.c 2028]
ERROR => DpWPCheck: W0 (pid 15634) died (severity=0, status=65280) [dpxxdisp.c 15663]
Y have stoped the instance 00 and then the shado instance (01) started OK and the process continue. but i can't get the instance 00 to go now.
I have set the projects parameters, and the parameter em/initial_size_MB is set to 1024 in both instance parameters.
Please help
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.