- Community Home
- >
- Servers and Operating Systems
- >
- Operating Systems
- >
- Operating System - HP-UX
- >
- Cluster files check
Categories
Company
Local Language
Forums
Discussions
Forums
- Data Protection and Retention
- Entry Storage Systems
- Legacy
- Midrange and Enterprise Storage
- Storage Networking
- HPE Nimble Storage
Discussions
Forums
Discussions
Discussions
Discussions
Forums
Discussions
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
- BladeSystem Infrastructure and Application Solutions
- Appliance Servers
- Alpha Servers
- BackOffice Products
- Internet Products
- HPE 9000 and HPE e3000 Servers
- Networking
- Netservers
- Secure OS Software for Linux
- Server Management (Insight Manager 7)
- Windows Server 2003
- Operating System - Tru64 Unix
- ProLiant Deployment and Provisioning
- Linux-Based Community / Regional
- Microsoft System Center Integration
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Community
Resources
Forums
Blogs
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
03-13-2003 05:23 PM
03-13-2003 05:23 PM
I have a two node(2*N4000/HP-UX 11.00/MCSG) cluster with total 6 packages running. 3 on each node.
The 6 different packages belongs to 6 different owners and hence it has been seen that these owners makes some changes while working on one node based on their requirement and is not updated in the 2nd node.
Now when one node fails and the package comes on the 2nd node it has been found that either the package will fail to start(say a new lvol has been added) or some of the application may not work(say entry has not been made in the /etc/services or cron.allow).
My question is:
Is there a script or a way through which I can check whether if any changes has been made in active node so that I can update the file/files in the standby node too respective to the package?
Thanks,
Sanjiv.
Solved! Go to Solution.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
03-13-2003 05:33 PM
03-13-2003 05:33 PM
Re: Cluster files check
There are a couple of ways around this. One way is to take the filesystem where they are making changes and put it into the package. If that won't work, you could write a script to check for file changes and copy them to the failover node, or even just copy them on a daily basis.
We use a script that provides daily information about our systems and it does a 'diff' on certain files and output of commands [vgdisplay -v, swlist -l product, etc.] between the current date and the previous date. If anything has changed it sends us an e-mail.
JP
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
03-13-2003 05:39 PM
03-13-2003 05:39 PM
Re: Cluster files check
Thanks for your advice.
Can you give me the script?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
03-13-2003 05:56 PM
03-13-2003 05:56 PM
Re: Cluster files check
The gist of it is that you have a text file with the previous days file you are checking, so you rename that and get a current copy of the file. Then, you do a 'diff' on the two files and send any output as e-mail to yourself. It's even handy for a shop with several Unix admins who don't always communicate what they are working on between themselves.
JP
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
03-13-2003 05:58 PM
03-13-2003 05:58 PM
Re: Cluster files check
Thanks again. I will be waiting for the script.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
03-19-2003 05:46 AM
03-19-2003 05:46 AM
Re: Cluster files check
I use rdist to keep these things in sync (it needs .rhosts files, but I'm guessing you have those anyway in a SG cluster)
rdist is very easy to configure, simply create a configuration file on your master which details the hosts to distribute to, and the files to distribute in the format:
HOSTS = ( node1 node2
node3 ... nodeN )
FILES = ( file1 file2
File3 ... fileN )
${FILES} -> ${HOSTS}
install -R ;
e.g. We have a three node cluster with nodes dilbert, dogbert and ratbert, we have nominated dilbert as our master. We have one package called oracle with three files that must be distributed (/etc/cmcluster/oracle/oracle.cntl, /etc/cmcluster/oracle/oramon.sh, and /etc/cmcluster/oracle/sqlnetmon.sh). Our rdistfile is installed as dilbert:/etc/cmcluster/rdist.cfg, and contains the following:
HOSTS = ( dilbert dogbert ratbert )
FILES = ( /etc/cmcluster/oracle/oracle.cntl
/etc/cmcluster/oracle/oramon.sh
/etc/cmcluster/oracle/sqlnetmon.sh
)
${FILES} -> ${HOSTS}
install -R ;
Now whenever we make any change to our package files on dilbert, we can distribute these changes to the other servers using the command:
rdist ???f /etc/cmcluster/rdist.cfg
If I am controlling the cluster I just do this manually... if others are responsible for it I might consider popping the above command in roots crontab on my 'master' host.
HTH
Duncan
I am an HPE Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
04-01-2003 01:15 AM
04-01-2003 01:15 AM
Re: Cluster files check
rdist seems to be good.
But I still prefer to have a script which can help me to find the difference of the important files between there two nodes?
Can someone give me or help me to write this script?
Thanks,
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
04-01-2003 01:27 AM
04-01-2003 01:27 AM
Re: Cluster files check
e.g.
compare file /etc/cmcluster/pkg1/pkg1.cntl on two hosts to local file:
file_to_diff=/etc/cmcluster/pkg1/pkg1.cntl
for node in node2 node3
do
echo "Differences on ${node}"
remsh ${node} "cat ${file_to_diff}" | diff ${file_to_diff} -
done
Hope this helps some...
Regards,
Duncan
I am an HPE Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
04-03-2003 12:01 PM
04-03-2003 12:01 PM
SolutionCluster Consistency Monitor
It details changes in your environment and is entirely customizable...
Rgds...Geoff