Openfiler 2.99 Active/Passive With Corosync, Pacemaker And DRBD

Want to support HowtoForge? Become a subscriber!
 
Submitted by wayner (Contact Author) (Forums) on Wed, 2011-04-27 17:34. :: Linux | Storage

Openfiler 2.99 Active/Passive With Corosync, Pacemaker And DRBD

Openfiler is a Linux based NAS/SAN Application which can deliver storage over nfs/smb/iscsi and ftp. It has a web interface over that you can control these services. This howto is based on the latest version of openfiler at this date, you can download it from the official homepage www.openfiler.com.

Thanks to the Openfiler team that made this howto possible.

 

1. Create Systems with following setup:

  • hostname: filer01
  • eth0: 10.10.11.101
  • eth1: 10.10.50.101
  • 500MB Meta partition
  • 4GB+ Data partition

 

  • hostname: filer02
  • eth0: 10.10.11.102
  • eth1: 10.10.50.102
  • 500MB Meta partition
  • 4GB+ Data partition

virtualip: 10.10.11.105 ( don't use on any adapter, we will make this later with corosync )

 

1.1 Create hosts file for easier access

root@filer01 ~# nano /etc/hosts

Add:

10.10.50.102	filer02

root@filer01 ~# nano /etc/hosts

On filer02 add:

10.10.50.101	filer01

 

1.2 Create/Exchange SSH Keys for easier file exchange

root@filer01 ~# ssh-keygen -t dsa

Generating public/private dsa key pair.
Enter file in which to save the key (/root/.ssh/id_dsa):
Created directory '/root/.ssh'.
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /root/.ssh/id_dsa.
Your public key has been saved in /root/.ssh/id_dsa.pub.
The key fingerprint is:

Do the same on filer02.

root@filer02 ~# ssh-keygen -t dsa

Then exchange the files:

root@filer01 ~# scp ~/.ssh/id_dsa.pub root@filer02:~/.ssh/authorized_keys

root@filer02 ~# scp ~/.ssh/id_dsa.pub root@filer01:~/.ssh/authorized_keys

And now you can exchange files between the nodes without entering a password.

 

2. Create meta/data Partition on both filers

Before we can actually start the cluster we have to prepaire both systems and let the data and meta partition sync before it can be used by corosync/pacemaker as the first cluster config will start drbd and take over the control of this service. So we prepaire our partitions this time before we do the actual cluster configuration as we did in openfiler 2.3.

 

2.1 Create DRBD Setup

Edit /etc/drbd.conf on filer01 and filer02:

# You can find an example in  /usr/share/doc/drbd.../drbd.conf.example
include "drbd.d/global_common.conf";
#include "drbd.d/*.res";
resource meta {
 on filer01 {
  device /dev/drbd0;
  disk /dev/sdb1;
  address 10.10.50.101:7788;
  meta-disk internal;
 }
 on filer02 {
  device /dev/drbd0;
  disk /dev/sdb1;
  address 10.10.50.102:7788;
  meta-disk internal;
 }
}
resource data {
 on filer01 {
  device /dev/drbd1;
  disk /dev/sdb2;
  address 10.10.50.101:7789;
  meta-disk internal;
 }
 on filer02 {
  device /dev/drbd1;
  disk /dev/sdb2;
  address 10.10.50.102:7789;
  meta-disk internal;
 }
}

After that create the meta-data on it, if you get errors when this happens, please empty out the filesystem with, if you have anything in /etc/fstab related to the partitions /meta then remove these lines. ( This happens when you create the meta partitions in the installation phase ).

dd if=/dev/zero of=/dev/drbdX

root@filer01 ~# drbdadm create-md meta
root@filer01 ~# drbdadm create-md data

root@filer02 ~# drbdadm create-md meta
root@filer02 ~# drbdadm create-md data

Now you can start up drbd with:

service drbd start

on both nodes.

Make one node primary:

root@filer01 ~# drbdsetup /dev/drbd0 primary -o
root@filer01 ~# drbdsetup /dev/drbd1 primary -o

 

2.2 Prepare the Configuration Partition

root@filer01 ~# mkfs.ext3 /dev/drbd0

root@filer01 ~# service openfiler stop

 

2.2.1 Openfiler to meta-Partition

root@filer01 ~# mkdir /meta
root@filer01 ~# mount /dev/drbd0 /meta
root@filer01 ~# mv /opt/openfiler/ /opt/openfiler.local
root@filer01 ~# mkdir /meta/opt
root@filer01 ~# cp -a /opt/openfiler.local /meta/opt/openfiler
root@filer01 ~# ln -s /meta/opt/openfiler /opt/openfiler
root@filer01 ~# rm /meta/opt/openfiler/sbin/openfiler
root@filer01 ~# ln -s /usr/sbin/httpd /meta/opt/openfiler/sbin/openfiler
root@filer01 ~# rm /meta/opt/openfiler/etc/rsync.xml
root@filer01 ~# ln -s /opt/openfiler.local/etc/rsync.xml /meta/opt/openfiler/etc/
root@filer01 ~# mkdir -p /meta/etc/httpd/conf.d

 

2.2.2 Samba/NFS/ISCSI/PROFTPD Configuration Files to Meta Partition

root@filer01 ~# service nfslock stop
root@filer01 ~# umount -a -t rpc-pipefs
root@filer01 ~# mkdir /meta/etc
root@filer01 ~# mv /etc/samba/ /meta/etc/
root@filer01 ~# ln -s /meta/etc/samba/ /etc/samba
root@filer01 ~# mkdir -p /meta/var/spool
root@filer01 ~# mv /var/spool/samba/ /meta/var/spool/
root@filer01 ~# ln -s /meta/var/spool/samba/ /var/spool/samba
root@filer01 ~# mkdir -p /meta/var/lib
root@filer01 ~# mv /var/lib/nfs/ /meta/var/lib/
root@filer01 ~# ln -s /meta/var/lib/nfs/ /var/lib/nfs
root@filer01 ~# mv /etc/exports /meta/etc/
root@filer01 ~# ln -s /meta/etc/exports /etc/exports
root@filer01 ~# mv /etc/ietd.conf /meta/etc/
root@filer01 ~# ln -s /meta/etc/ietd.conf /etc/ietd.conf
root@filer01 ~# mv /etc/initiators.allow /meta/etc/
root@filer01 ~# ln -s /meta/etc/initiators.allow /etc/initiators.allow
root@filer01 ~# mv /etc/initiators.deny /meta/etc/
root@filer01 ~# ln -s /meta/etc/initiators.deny /etc/initiators.deny
root@filer01 ~# mv /etc/proftpd /meta/etc/
root@filer01 ~# ln -s /meta/etc/proftpd/ /etc/proftpd

 

2.2.3 httpd Modules for Openfiler

root@filer01 ~# rm /opt/openfiler/etc/httpd/modules
root@filer01 ~# ln -s /usr/lib64/httpd/modules /opt/openfiler/etc/httpd/modules

Now do a start and see if Openfiler can run:

root@filer01 ~# service openfiler start

 

2.2.4 filer02 Openfiler Configuration

root@filer02 ~# service openfiler stop
root@filer02 ~# mkdir /meta
root@filer02 ~# mv /opt/openfiler/ /opt/openfiler.local
root@filer02 ~# ln -s /meta/opt/openfiler /opt/openfiler

 

2.2.5 Samba/NFS/ISCSI/ProFTPD Configuration Files to Meta Partition

root@filer02 ~# service nfslock stop
root@filer02 ~# umount -a -t rpc-pipefs
root@filer02 ~# rm -rf /etc/samba/
root@filer02 ~# ln -s /meta/etc/samba/ /etc/samba
root@filer02 ~# rm -rf /var/spool/samba/
root@filer02 ~# ln -s /meta/var/spool/samba/ /var/spool/samba
root@filer02 ~# rm -rf /var/lib/nfs/
root@filer02 ~# ln -s /meta/var/lib/nfs/ /var/lib/nfs
root@filer02 ~# rm -rf /etc/exports
root@filer02 ~# ln -s /meta/etc/exports /etc/exports
root@filer02 ~# rm /etc/ietd.conf
root@filer02 ~# ln -s /meta/etc/ietd.conf /etc/ietd.conf
root@filer02 ~# rm /etc/initiators.allow
root@filer02 ~# ln -s /meta/etc/initiators.allow /etc/initiators.allow
root@filer02 ~# rm /etc/initiators.deny
root@filer02 ~# ln -s /meta/etc/initiators.deny /etc/initiators.deny
root@filer02 ~# rm -rf /etc/proftpd
root@filer02 ~# ln -s /meta/etc/proftpd/ /etc/proftpd

 

2.3 Prepare the Data Partition

Change the lvm filter in the

/etc/lvm/lvm.conf

file from:

filter = [ "a/.*/" ]

to

filter = [ "a|drbd[0-9]|", "r|.*|" ]

Exchange this file to the other filer node

root@filer01 ~# scp /etc/lvm/lvm.conf root@filer02:/etc/lvm/lvm.conf

After that we can create the actual used stuff:

root@filer01 ~# pvcreate /dev/drbd1
root@filer01 ~# vgcreate data /dev/drbd1
root@filer01 ~# lvcreate -L 400M -n filer data

 

3. Start Corosync and create a configuration for it:

3.1 Create Corosync authkey

root@filer01~# corosync-keygen

( Press the real keyboard instead of pressing keys in an ssh terminal. )

Copy the authkey file to the other node and change the fileaccess:

root@filer01~# scp /etc/corosync/authkey root@filer02:/etc/corosync/authkey
root@filer02~# chmod 400 /etc/corosync/authkey

 

3.2 Create a file named pcmk /etc/corosync/service.d/pcmk

root@filer01~# vi /etc/corosync/service.d/pcmk

service {
        # Load the Pacemaker Cluster Resource Manager
        name: pacemaker
        ver:  0
 }

 

3.2.1 Exchange this file to the other node

root@filer01~# scp /etc/corosync/service.d/pcmk root@filer02:/etc/corosync/service.d/pcmk

 

3.3 Create the corosync.conf file and change it to present your lan net ( bindnetaddr )

root@filer01~# vi /etc/corosync/corosync.conf

# Please read the corosync.conf.5 manual page
compatibility: whitetank
totem {
        version: 2
        secauth: off
        threads: 0
        interface {
                ringnumber: 0
                bindnetaddr: 10.10.50.0
                mcastaddr: 226.94.1.1
                mcastport: 5405
                ttl: 1
        }
}
logging {
        fileline: off
        to_stderr: no
        to_logfile: yes
        to_syslog: yes
        logfile: /var/log/cluster/corosync.log
        debug: off
        timestamp: on
        logger_subsys {
                subsys: AMF
                debug: off
        }
}
amf {
        mode: disabled
		}

 

3.3.1 Exchange the file to the other node

root@filer01~# scp /etc/corosync/corosync.conf root@filer02:/etc/corosync/corosync.conf


Please do not use the comment function to ask for help! If you need help, please use our forum.
Comments will be published after administrator approval.
Submitted by Anonymous (not registered) on Fri, 2013-08-09 21:33.
Wouldn't it make more sense to use drbdlinks instead of moving things into the meta directory manually?
Submitted by Sander (not registered) on Wed, 2012-08-22 08:48.

Thank you for this info! It really helped me. I had still an issue with the failover part caused by a small bug in the iscsi package of openfiler. This thread helped me solving it: https://forums.openfiler.com/index.php?/topic/5739-lvm-cluster-resource-migration-issue/

Submitted by Anonymous (not registered) on Fri, 2012-03-02 20:12.
Can someone please tell me how to set up a 3rd offsite node in this scenario?
Submitted by Eugeny (not registered) on Mon, 2012-01-16 13:42.
The command:

lvcreate -L 400M -n filer data

Why "-L 400M"? If this is data volume, why not to use all 4GB?

Submitted by Anonymous (not registered) on Wed, 2012-05-30 18:37.
This is just the initial start volume to get things going. It doesn't matter what size you put it. You can create any volume size later with the web interface after everything is completely setup.
Submitted by Anonymous (not registered) on Tue, 2011-06-28 07:59.

 I got following:

[root@filer02 ~]# crm configure show
node filer01 \
        attributes standby="on"
node filer02 \
        attributes standby="on"

 

 How to change to attributes standby="off" ?

Submitted by Saiful (not registered) on Wed, 2011-12-21 12:19.

You may edit those item:

 # crm configure

 crm(live)configure# edit filler01

Submitted by SamTzu (registered user) on Sat, 2011-05-07 17:15.

This here is a good reason to change to FreeNAS. I did several years ago.
It's way easier to install and use.

Sam

Submitted by Anonymous (not registered) on Sat, 2012-03-03 02:44.
I really am having a hard time considering FreeNAS as a SAN solution for HA iSCSI deployment.  Sorry, but I think you must be referring to some home media server or something.  This is all about grown up stuff - not toys.
Submitted by wayner (registered user) on Mon, 2011-05-09 10:52.

Because its so easy to setup FreeNAS as High Available Solution? I doubt so ... but why dont you write a howto about it because i had a hardtime using freenas after all.