Separate configuration directory?

Is it a common practice to store the directory
configuration on another directory server? I seem to
have more reasons to do this now.
If you use a separate directory for directory
configuration, how do you do failover? I could set up
2 configuration servers and enable replication but if
all my servers are pointed to one and it fails, how
do you get them to point to the second server? Is
there a way to have automatic failover? Is there a
way to manually set the servers to point to the
second one?We have two instances on the same directory server, one for the config info (slapd-config) and the other for our actual data (slapd-data). We have four servers, two masters, two consumers, set up in this manner. The config info is not replicated between the the servers, while our actual data is. It is not recommended that you replicate the configuration information between different directory servers. It is a bit of a pain to have to make changes on 4 servers, but it is not overly onerous. If you have more servers such that making changes would take a long time, you could figure out the ldif needed to make the change and script it rather than using the console.
If you were to have a single config directory, rather than failover, create a multi-master replica and put it behind a load balancer. The load balancer should be able to do a helath chack and send requests to an active host.

Similar Messages

IDS 5.x: Is a separate configuration directory server instance even needed?

Because iDS 5.x supports multiple database instances, why is there a need to suggest installing the directory server with a separate configuration directory server instance (on say port 390)?
In large multi-tiered iDS 4.x directory server implementations, using a single master configuration directory server is very cumbersome, so many implementations are using a local configuration directory server instance, on port 390 with the data instance on the usual port 389.
With the iDS 5.x release, we will be implementing local directory server instances, all on port 389, on all of our tiered LDAP servers. There will be no separate configuration directory server instance, as it is not needed.
At this point, I am questioning the "best practices" suggested by the SunOne documentation to use a separate configuration directory server instance. It does not need to be on a separate port now with iDS 5.x supporting multiple databases.
And I note that the present "/usr/sbin/directoryserver setup" script will not allow for a directory instance to be installed on port 390.
Adam

Yes, with iDS 5.X, it's much safer to have the o=NetscapeRoot tree served by the same instance as your corp/user tree. The separation approach was created to avoid the pitfalls when importing LDIF, but the -n argument to ldif2db lets you avoid such problems.

Purpose of Separate Configuration Directory

I know that Sun best practices recommend that the configuration directory instance be separated (placed on a different physical server) in a directory environment.
Other than the ability to manage all directories in a single administration domain, what does this accomplish?
Is there anything in a replicated multi-master environment that would require such a setup, with all master and consumer instances pointing to the same configuration directory?
Is it possible to set up a replicated multi-master environment with configuration directories on each individual server?
When do "user" directories need to access the configuration directory instance, other than during initial setup? If the configuration directory is on the WAN relative to a user directory, could inability to access the config directory prevent the user directory from starting/stopping/accepting requests/otherwise functioning?
Thanks in advance.

I don't disagree, however I think you can go too far. If everything is in one o=netscapeRoot, it's like putting all your eggs in one basket. If you loose that directory, you loose console access on all SunOne instances. And while having a multi-master config directory aleviates the problem somewhat, usually loosing the config directory does not occurr because of a hardware problem - rather user error or an upgrade gone bad - and the error usually replicates just fine. Backups are key before any operations that changes or could potentially chage the config instance. Personally I like to have one config instance per <Server Root>. It does not take much in resources.

Use of configuration directory server

Maybe this is a dumb question but since I dont know the answer:
Does one need to buy a production license of the Directory Server to use as a configuration directory server?
I will have six web servers in my production environment. But only one or two admin user accounts for secure access.
Thank,
-amit

Both 2005Q1 and 2005Q4 are considered 6.2. I would
use 2005Q4. Actually, I DID use 2005Q4.
Why use bits that are already superceeded? Is it the
cost of 4 CD's?So, can this happen in two separate stages?:
1. First upgrade all directory servers to 5.2p4 and migrate the current messaging server 5.2 configuration.
2. At a later date, once directory services have been upgraded and are stable, the upgrade of the messaging server can be done.
Thanks

Automake with a separate include directory (also curses)

I am building a program, and I am trying (unsuccessfully) to add autotools support to it, for easier distribution. However, because I chose to have a separate include/ directory for my header files, I am encountering innumerable difficulties. I have a top level Makefile.am, and another in src/, and this one should be the one that does the including, I think. Even if I specify the header files in sources, they are not included. I do know about AM_CPPFLAGS, but somehow, this seems like the wrong thing to do.
I guess what I'm really wondering is there any way to include header files from a separate directory without using AM_CPPFLAGS="$AM_CPPFLAGS -I../include", which seems like it won't include the sources in the distribution, according to what I have read.
A separate problem I am having is that I am not able to include libraries with my program. It depends on ncurses, so how do I specify the -lncurses option? Again with AM_CPPFLAGS? I would rather not.

Maybe my old post will help, or at least move you along a bit: https://bbs.archlinux.org/viewtopic.php … 2#p1255512
Specifically, see my "configure.ac" file, the section for "AC_CHECK_LIB" is used to include a library (like ncurses).

ERROR: Cannot start mail services, configuration directory does not exist

I'm running an Xserve G5 with one internal boot drive and one RAID using the other two internal drives. I used Apple's Disk Utility to set-up the RAID.
Back in early December, I moved my mail database and mail store to the RAID. Ever since, when I restart the server, the mailaccess.log generates the following error:
Jan 23 18:51:33 mail master[47]: ERROR: Cannot start mail services, configuration directory does not exist: /Volumes/RAID/mail.data/db
I'm speculating this is a timing error that gets created when Postfix initially can't find the RAID. Then, it must try again and find it because the error message does not repeat and mail starts running.
Is this something to worry about?

Hi Joel,
You're right, it's a timing issues. The mail system is ready for work before the file system is mounted. It's benign.
Jeff

Error configure directory server 5.2

Hi,
I'm getting this error when doing issuing
bash-3.00# directoryserver configure
after installing Directory Server with Configure Later option
[slapd-elara]: starting up server ...
[slapd-elara]: [29/Aug/2009:10:40:37 +0100] - Sun Java(TM) System Directory
Server/5.2_Patch_4 B2005.230.0041 (64-bit) starting up
[slapd-elara]: [29/Aug/2009:10:40:38 +0100] - Listening on all interfaces port
389 for LDAP requests
[slapd-elara]: [29/Aug/2009:10:40:38 +0100] - slapd started.
Your new directory server has been started.
error: can't bind to server:Unable to bind to server. (Invalid credentials (49)
returned from ldap_simple_bind_s(cn=Directory Manager))
Could not configure server.
Configuration of the Directory Server failed.
Error Directory Server configuration failure
Error Configuration of the server(s) failed.
Can someone give me a hint?
Best regards

Problem solved.
1 - uninstalled Directory Server
/var/sadm/prod/SUNWds-entsys4/uninstall
2 - removed all the packages used in a manual instalation
for i in SUNWasha SUNWasvc SUNWasvcp SUNWasvr SUNWasvu SUNWdsha SUNWdsvcp SUNWdsvh SUNWdsvhx SUNWdsvpl SUNWdsvr SUNWdsvu SUNWdsvx SUNWicu SUNWicux SUNWjss SUNWldk SUNWldkx SUNWpr SUNWprx SUNWsasl SUNWsaslx SUNWtls SUNWtlsx; do echo $i;pkgrm $i;done
3 - re-install it with configure later option
java_es_05Q4_directory/Solaris_sparc/installer
4 - Configure Directory Server
directoryserver -u 5.2 configure
5 - Configure Administration Server
mpsadmserver configure
And everything started working.

Change of configuration Directory programmatic ...

Hello,
we are searching for a way to change the TestStand configuration Directory programmatically ...
Before a sequence will be loaded... (maybe within the sequence file callback - SequenceFileLoad - if possible)
(Menue: Configure>Station Options>Preferences [Configuration Directory])
Is the TestStand API a workable way? ... Maybe someone of you have a nice example...
Greetings from Germany
Norbert

Hello Norbert,
we're doing that by a registry file before we open Teststand.
Create a test file, enter
Windows Registry Editor Version 5.00
[HKEY_LOCAL_MACHINE\SOFTWARE\National Instruments\TestStand\3.5]
"CfgLocation"="c:\\myProject\\myConfigLocation"
and save it with extention ".reg". On a double-click, windows sets the config location to C:\myconfigLocation. See attached file.
By the way, if the directory doesn't exist, Teststand creates it on startup. But if the drive doesn't exist, it refuses to start with ugly error messages.
But as mentioned, we switch the location when Teststand is closed.
Schöne Grüße vom Bodensee
Thomas
Attachments:
SetTs35CfgPath.zip ‏1 KB

Move Configuration directory server

We are in the process of transitioning from DS5.2 to DS6.3.1. A few of the consumers will be remaining at 5.2 level for some time (budget reasons). Currently they all have their configuration stored on one of our 5.2 masters. When the time comes to take down this old master we'd like to move the configuration directory (the o=NetscapeRoot information) to the individual consumers. Anyone know how to get the config for a particular consumer, upload it (to a file or a newly created o=netscaproot on the consumer), and the configure the admin server to look for it in the new location?
Thanks for any help.

Thanks,
Yes, I do think it will be better to take the whole NetscapeRoot tree instead of trying to get just the host specific data now that I think about it. But the other question of getting the admin servers on the consumers to look for their info in the new location - I tried editing {install-root}/admin-serv/config/adm.conf and changing the 'ldaphost' line to a new host, but even after restart it kept going to the old server. Any ideas on getting this to change?

Using version 4 as configuration directory for 5

Is there any issue with using and instance of version 4.x as the
configuration directory for version 5?
Thx
-Wajih

As far as I know there is no issue with doing this.
Ludovic

Set TestStand Configuration Directory

HI,
I'm trying to set TestStand Cfg directory from a LabVIEW OI (I customized the full featured OI shipped with TestStand). This works fine only if I restart TestStand. I understand I only need to shutdown the Engine. My question is how can I restart the Engine without restarting everything? (there is no such thing as a Start Engine method) Could a GetEngine method work? What would be the best way to accomplish this task? I don't want my customer to restart the software.
I need to set the configuration directory because different softwares might be installed on the same station with different configurations (different station globals for example), or maybe developping 2 applications for 2 clients on the same computer.
Ps: I also posted this question on LAVA: http://lavag.org/topic/14439-configuring-teststand-cfg-directory/
Thank you all in advance,
Regards,
Rodéric L
Certified LabVIEW Architect
Solved!
Go to Solution.

Hello Rodéric,
                   i want to set Cfg folder path through operator interface.vi so I've placed the following program into operator interface.vi before intitalize.vi (the operator interface.vi is located @ C:\Programm files\National Instruments\Testsand2010\opeartor Interface\LabVIEW.....)
            The Teststand maneger reference from operator interface.vi is passes through this vi, as i run the operator interface.vi this program executes without error but i get error at the next property node "IApplicationMgr --> Start" which is placed into intialize.vi
Error : Two Application maneger is opened
          I have tried two options, for each i get same error, the options are as follows
1) The TSUI.IApplication maneger reference is opened and closed in operator interface as well as the above program seperately
2) I have passed only one TSUI.IApplication maneger reference through out operator interface
Thanks & Regards
Niranjan Kamble
CLAD/Teststand

Does iDAR 5 need a configuration directory?

Does iDAR 5 need a configuration directory? It looks like it requires one to
store its configuration and it is iDS 5. iDAR 2.1 did not need to store its
confuration into a directory somwhere.

John Miller wrote:
Does iDAR 5 need a configuration directory? It looks like it requires one to
store its configuration and it is iDS 5. iDAR 2.1 did not need to store its
confuration into a directory somwhere.iDAR 5 does require a configuration directory, it is fully integrated with the common iPlanet admin server that uses the directory to store all the configuration information.
Daniel

EMCLI default configuration directory

How do I find the default configuration directory for EMCLI on the OMS host? I changed the EM Console ports and need to re-configure EMCLI to use the new port. I can get the configuration using:
$ $OMS_HOME/bin/emcli setup
Oracle Enterprise Manager Cloud Control 12c Release 2.
Copyright (c) 1996, 2012 Oracle Corporation and/or its affiliates. All rights reserved.
Instance Home : /u01/app/oracle/Middleware/gc_inst/em/EMGC_OMS1/sysman/emcli/setup/.emcli
Verb Jars Home : /u01/app/oracle/Middleware/gc_inst/em/EMGC_OMS1/sysman/emcli/setup/.emcli
EM URL : https://oms1.example.com:7799/em
EM user : sysman
Trust all certificates : true
Auto login : false
But it doesn't display the configuration directory. I'd like to re-use the existing directory.
Thanks,
Leighton

Hello Leighton,
have you tried to change this using:
./emcli setup -url=http://myworkstation.us.oracle.com:em_port/em -username=em_user
Regards
Rob
http://oemgc.wordpress.com

Global Configuration Directory

I have installed the latest version (1.5.1 build 5440) on Ubuntu 8.04 64 bit.
If i launch the application from the terminal it allows me to set up connections from the tnsnames.ora file.
If I launch it from the menu it does not. It appears that my environment variables (such as TNS_ADMIN) are not used when I launch the application from the menu.
The User Guide refers to the Global Configuration Directory but does not say where or what it is.
Please point me in the right direction.
Regards;
Robert

Chris;
Yes all my environmental variables are set in .bashrc. I created a .bash_profile and even a .profile with all the relevant Oracle environmental variables. There is no difference.
Is there a file like <sqldeveloper_home>/sqldeveloper/bin/sqldeveloper.conf where I can set the TNS_ADMIN path in the same way that I set the JAVA_PATH?
Can I set it in the <sqldeveloper_home>/sqldeveloper/bin/sqldeveloper.conf file, and if so what would I call it?
I have tried setting TNS_ADMIN in the sqldeveloper.conf file, it did not work.
Regards;
Robert

Replicating Configuration Directory

Hello,
I have replicated my config directory (o=NetscapeRoot) from one server (ldap1) to a second server(ldap2). My question is, what do I need to change on my directory servers (ldap2, ldap3, ldap4) to use the configuration directory on ldap2 if ldap1 goes down?
I am running DS5.2.
I appreciate any help you may be able to give.
Thanks,
Eric

So you have 2 config dirs. You import the one into the other and then replicate that back to the other one you just dumped. Now you have what you need.
We use an alteon by nortel. Both configs run on the same port but they dont have to. You set a virtual host for the port up on the alteon in failover mode to direct all traffic to one and fail over to the other if the first one is down. I then went and set my actual directory servers to look at that virtual.

Separate configuration directory?

Similar Messages

Maybe you are looking for