Working with disk sets from replicated LUNs

Thread Tools Search this Thread
Operating Systems Solaris Working with disk sets from replicated LUNs
# 1  
Old 03-17-2014
Working with disk sets from replicated LUNs

Hi everybody
We have a Sun Cluster with two nodes, connected to a number of SAN disks over fiber cables. We made SVM disk sets on those disks for our application needs.
Now we constructed another site in another metropolitan area, but with only one node (no cluster), and connected it to the main site through WAN. We replicated the LUNs from the main site to the secondary site.
Now the question is: Can I (and how can I) work with the replicated LUNs so that I can clone the application from the main site to the secondary site and let it see the disks sets, soft partitions...etc that were replicated?
# 2  
Old 03-17-2014
You have the option of a geographic cluster - remote site + two local nodes.
Consider adding additional inexpensive (x86-based) "quorum servers" so you can have cluster failover.

Oracle Solaris Cluster Geographic Edition Overview - Oracle Solaris Cluster Geographic Edition Overview
# 3  
Old 03-18-2014

Thanks for you reply. I am not talking about a planned architecture; I am talking about an installed and working one that I inherited. Changing the cluster type in the time being is not an option.
Let me rephrase: I need to see the up and running disksets that exist in the cluster in the DC in the same manner (same device names, data..etc) in a single node in another location.
I already tried metaimport -r -v but it returns "no unconfigured devices detected". What are my options? and why metaimport returns this message?

---------- Post updated at 08:22 AM ---------- Previous update was at 01:30 AM ----------

Ok guys I managed to solve it myself. For anybody reading this and looking for a solution here is the procedure:
(I am assuming that data has been replicated and that you do have read/write access to it)
Let's say you want to replicate disk set abc123 that has a number of volumes/soft partitions from machine main01 to machine dr01:
  1. You need to have a clear and accurate mapping between the devices you have in the main site and the devices that was replicated to you in the DR site.
  2. On the dr01 create a diskset with the same name as the original one on main01
    metaset -s abc123 -h dr01

  3. Start adding the replicated disks to the diskset in the same order as it was entered on main01 (you can determine this by running
    metastat -s abc123

    and see the order at the end.
  4. On main01
    metastat -s abc123 -p

    This command will print the SVM configuration. Take a copy of all the output.
  5. For every volume in the diskset you will see a line like the following
    abc123/d4000 1 6 /dev/rdsk/c2t50000975F007B949d25s0 /dev/rdsk/c2t50000975F007B949d24s0 /dev/rdsk/c2t50000975F007B949d26s0 /dev/rdsk/c2t50000975F007B949d27s0 /dev/rdsk/c2t50000975F007B949d28s0 /dev/rdsk/c2t50000975F007B949d29s0 -i 32b

    you need to replace those devices with the corresponding replicated devices. Again the order is very important.
  6. If everything is done correctly you can issue the following command
    metainit -s abc123 -a

    which will initialize the diskset with all the required volumes and soft partitions.
If this option fails, you can add the volumes manually and use the metarecover command to recover the soft partitions.
Login or Register to Ask a Question

Previous Thread | Next Thread

5 More Discussions You Might Find Interesting

1. Shell Programming and Scripting

Disk Space Utilization in HTML format working in one environment and not working on the other

Hi Team, I have written the shell script which returns the result of the disk space filesystems which has crossed the threshold limit in HTML Format. Below mentioned is the script which worked perfectly on QA system. df -h | awk -v host=`hostname` ' BEGIN { print "<table border="4"... (13 Replies)
Discussion started by: Harihsun
13 Replies

2. Shell Programming and Scripting

Format data by consolidating replicated obs

Please assist with the following. I have readings R1, R2,..Rn from plants from different fields , sometimes replicated. I need to first consolidate the replicates by their mean value per plant per field. Then I need to reformat the data by partially transposing it. The missing values... (3 Replies)
Discussion started by: jianp83
3 Replies

3. Shell Programming and Scripting

Preventing script from being replicated on a defined number of hosts

ok. i have an extensive program written predominantly in borne shell. i have to give an "evaluation" copy of this program to a user so she can test it out and see if she wants it. problem is, i dont have an evaluation copy. and even if i did, im worried the evaluation copy can be edited to... (8 Replies)
Discussion started by: SkySmart
8 Replies

4. Shell Programming and Scripting

How to replicated records using sed

I need to generate exactly 102 duplicates in a file using sed. Here's an example of my input: Grid-ref = 1, 148 Grid-ref = 1, 311 Grid-ref = 1, 428 I need: Grid-ref = 1, 148 Grid-ref = 1, 148 Grid-ref = 1, 148 Grid-ref = 1, 148 ... repeated 102 times, then Grid-ref = 1, 311... (2 Replies)
Discussion started by: Russ
2 Replies

5. Virtualization and Cloud Computing

Clouds (Partially Order Sets) - Streams (Linearly Ordered Sets) - Part 2

timbass Sat, 28 Jul 2007 10:07:53 +0000 Originally posted in Yahoo! CEP-Interest Here is my follow-up note on posets (partially ordered sets) and tosets (totally or linearly ordered sets) as background set theory for event processing, and in particular CEP and ESP. In my last note, we... (0 Replies)
Discussion started by: Linux Bot
0 Replies
Login or Register to Ask a Question