07-07-2014 01:41 AM
First of all hi everybody,
I'm pretty new to Linux and Dataprotector.
So my task is to install 27 Cell Managers and Installation Servers (both on the same) on OEL 6.5 (similar to RHEL 6.5)
I've done that so far.
Now I've got the Problem that my Dataprotector services are not starting.
[root@server ~]# /opt/omni/sbin/omnisv status ProcName Status [PID] =============================== crs : Active  mmd : Active  kms : Active  hpdp-idb : Active  hpdp-idb-cp : Down hpdp-as : Down omnitrig : Active Sending of traps disabled. =============================== Status: At least one of the Data Protector processes/services is not running. [root@server ~]#
/opt/omni/sbin/omnisv stop works fine, but takes a while.
/opt/omni/sbin/omnisv start takes even longer and gives the error:
Cannot start "hpdp-idb-cp" service, system error:  Unknown error 1053
Hope somebody can tell me how to fix this error.
By the way I've got another error:
I can't connect with the GUI. It keeps asking me for a Username and a Password.
vi /etc/opt/omni/server/users/UserList "root" root * server.site admin "WebReporting" "java" "applet" webreporting admin "*" * * * admin
So I thought i just could configure it via CLI. I installed the CC Component.
I get the omnicc man page, but I can't type any other command like:
[root@server ~]# omnicc -check_licenses -bash: omnicc: command not found
All Servers got the same Issue.
HP Data Protector A.08.10: OMNICC, internal build 154, built on Wed 04 Dec 2013 07:15:39 AM WAT
Oracle Linux Server release 6.5
Kernel \r on an \m
Thanks everybody for further instructions
Sorry for the bad spelling.
07-07-2014 02:28 AM - edited 07-07-2014 02:43 AM
we have our dataprotector cell on oel 6.5 too.
We start& stop it with "service omni stop" | "service omni start". Maybe this will work.
If not, you may want to check the log files in /var/opt/omni/log for errors.
It will be probably the debug.log file
We had the same login error before. But you probably dont want to hear the solution :)
This is an error when you dont use an FQDN (full qualified domain name) while the installation of the cell manager. Unfortunatley you have to reinstall the cell manager with an FQDN. But maybe you should wait for other replies, if im wrong. To check it type in following command:
it has to be like "test.test.com". If its not... you have to change it and reinstall the cell manager
If your PATH_variable is not configured, you have to you your omni commands like this:
If you dont want to do this you can edit your .profile:
(We installed the dataprotector in /opt/omni , if you dont, replace the path in the following)
cd ~ #( Now edit the file with nano, or use "vi" instead of nano, but if #you´re new to linux i doubt you know vi, and its more difficult): nano .bashrc #Then add following: export PATH=$PATH:/opt/omni/bin/:/opt/omni/lbin/:/opt/omn
With ctrl + x you can save the file.
Then save it and logout and login. Then you can use the commands without the path.
07-07-2014 03:59 AM
Thank you for your reply,
We start& stop it with "service omni stop" | "service omni start"
This one works for me as well, but with the same error.
To the FQDN, i haven't installed these Servers.
I know I need FQDN and working DNS, but I'm only unsing the /etc/hosts for all hosts.
So there is not DNS, after I little check i recognised the misstake.
[root@servername ~]# hostname servername.site [root@servername ~]# hostname -f servername [root@servername ~]# domainname (none)
Somebody changed the FQDN and hostname field by misstake I think. So I have to correct it on every server right?
That will at least take a week to change everything. :'(
By the way:
Is it a Problem if hostname gives the output *servername*.site? Can I just put *servername*.site as FQDN? Or do I need
*servername*.site.cm.ng or something like that?
07-07-2014 04:14 AM
As per my knowledge,
DP takes the hostname during installation and this hostname will be associated with the Cell Manager(CM).
You have to provide this CM FQDN name while commecting using GUI. So this name must be pingable.
You can change /etc/hosts in one machine, associate a seperate ip and make it pingable. But how can you do this for all 27 machines when all of them have same hostname.
And also, I think instead of installing 27 CM and IS,
Install CM and IS in one machine.
Add other 26 machines as DP clients to the CM using the IS. This will be fast and you will have centralised management.
Do a POC on one machine and then repeat on all other systems.
Note: Firewalls must be disabled or properly configured. And please sync the time, to avoid weired SSL issues.