headers
Su Chen | 17 Jun 2013 20:16
Favicon

Why cman is started at rc0.d and rc6.d

Hi All,

I am very new to pacemaker, cororsync and cman. I installed the packages on an Ubuntu machine. ("aptitude
install pacemaker cman fence-agents")
To my surprise, cman has a link under rc0.d and rc6.d, why cman need to be started while system is shuting down?

root <at> SuTH3:/etc# ls -l /etc/rc0.d/S05cman /etc/rc6.d/S05cman
lrwxrwxrwx 1 root root 14 May 18 23:55 /etc/rc0.d/S05cman -> ../init.d/cman
lrwxrwxrwx 1 root root 14 May 18 23:55 /etc/rc6.d/S05cman -> ../init.d/cman

Thanks,
Su
_______________________________________________
Linux-HA mailing list
Linux-HA <at> lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
Permalink | Reply | 1 comment |
headers
Parkirat | 17 Jun 2013 20:12
Picon

Resource Collocation v/s Resource Groups

Hi,

Is there any difference between Resource Collocation and Resource Groups?

I grouped 2 resources both having migration_threshold=2 and
monitor_interval=60s. When, I stopped 1 of the resource from the group, I
did not restarted. However, when I was configuring the resource not in the
group, the resource started on manually stopping it.

Also is there any way to order the sequence of the resource in a group?

Regards,
Parkirat Singh Bagga.

--
View this message in context: http://linux-ha.996297.n3.nabble.com/Resource-Collocation-v-s-Resource-Groups-tp14699.html
Sent from the Linux-HA mailing list archive at Nabble.com.
_______________________________________________
Linux-HA mailing list
Linux-HA <at> lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
Permalink | Reply | 5 comments |
headers
Digimer | 17 Jun 2013 19:20
Picon

Re: Heartbeat haresources with IPv6

Ho Thiago,

   Heartbeat is deprecated and has not been developed in some time. 
There are no plans to restart development, either. It is _strongly_ 
advised that new setups use corosync + pacemaker. You can use the IPv6 
resource agents with it, too.

   The best place to look is on clusterlabs.org's "Cluster from Scratch" 
tutorial. It covers as the first example setting up an (IPv4) virtual IP 
address. It should be easy to adapt that to your IPv6 implementation. 
You will see two versions; One for crmsh and one for pcs. I would 
recommend the crmsh version for Ubuntu.

Cheers

On 06/17/2013 11:35 AM, listas <at> adminlinux.com.br wrote:
> Hi,
>
>
> I'm using Ubuntu 12.04 + Heartbeat 3.0.5-3ubuntu2 to provide high
> availability for some IP addresses.
> I want to configure an IPv6 address on my haresources. I did this:
>
> File /etc/heartbeat/haresources:
>
> server.domain.com \
> &nbsp;&nbsp;&nbsp; 192.168.2.62/32/eth1 \
> &nbsp;&nbsp;&nbsp; 192.168.2.64/32/eth1 \
> &nbsp;&nbsp;&nbsp; 192.168.2.72/32/eth1 \
> &nbsp;&nbsp;&nbsp; IPv6addr::2001:db8:38a5:8::2006/48/eth1 \
(Continue reading)

Permalink | Reply | 0 comments |
headers
listas | 17 Jun 2013 17:35
Picon

Heartbeat haresources with IPv6

Hi,

 I'm using Ubuntu 12.04 + Heartbeat 3.0.5-3ubuntu2 to provide high availability for some IP addresses.
 I want to configure an IPv6 address on my haresources. I did this:

 File /etc/heartbeat/haresources:

 server.domain.com \
 &nbsp;&nbsp;&nbsp; 192.168.2.62/32/eth1 \
 &nbsp;&nbsp;&nbsp; 192.168.2.64/32/eth1 \
 &nbsp;&nbsp;&nbsp; 192.168.2.72/32/eth1 \
 &nbsp;&nbsp;&nbsp; IPv6addr::2001:db8:38a5:8::2006/48/eth1 \
 &nbsp;&nbsp;&nbsp; MailTo::adm <at> domain.com

 The IPv4 addresses work fine, but I'm not getting success with the IPv6 address.
 My logs shows this message:
 ResourceManager[22129]: info: Running /etc/ha.d/resource.d/IPv6addr 2001:db8:38a5:8
2006/48/eth1 start
 ResourceManager[22129]: CRIT: Giving up resources due to failure of IPv6addr::2001:db8:38a5:8::2006/48/eth1
 ResourceManager[22129]: info: Running /etc/ha.d/resource.d/IPv6addr 2001:db8:38a5:8
2006/48/eth1 stop
 ResourceManager[22129]: info: Retrying failed stop operation [IPv6addr::2001:db8:38a5:8::2006/48/eth1]

 Apparently there is a conflict between the characters '::' inside the IPv6 address and the separator '::'
used in the haresources. But I would not like have to expand the IPv6 address. 

 Does anyone know a way to avoid this conflict?

 Thanks!
 --
(Continue reading)

Permalink | Reply | 0 comments |
headers
Sven Arnold | 16 Jun 2013 18:07
Picon

Trouble with multipath iSCSI and KVM live migration

Dear All,

I have set up a three node cluster with shared storage (DRBD 
active/passive) which exports iSCSI Volumes (TGT) containing KVM/QEMU 
disk images.

The iSCSI Target is configured as one resource and accessible on two 
floating ip addresses to allow multipath I/O for speed and redundancy.

The VM hosts are accessing the volumes via open-isci using dm-multipath 
(grouping_policy multibus).

While migrating the iSCSI Target from A to B everything works fine.
But if I try to live migrate a virtual machine I experience file system 
corruptions inside the virtual machine. So, somehow the switching of the 
iSCSI/Multipath Sessions is not handled properly by the VM hosts.

I have configured iSCSI timeouts rather short (noop_out_timeout 5 
seconds) and "no_path_retry queue" on the multipath device.

My question(s):

1) Is it conceptually wrong what I am trying to accomplish?

2) Is it valid to use "no_path_retry queue" in such a setup?

2) Did I miss some important configuration options (timings, etc.)?

3) Is TGT multipath capable?
(Continue reading)

Permalink | Reply | 0 comments |
headers
Lars Ellenberg | 15 Jun 2013 23:57
Favicon

Re: Heartbeat haresources with IPv6

On Fri, Jun 14, 2013 at 03:29:49PM -0300, listas <at> adminlinux.com.br wrote:
> Hi,
> 
> I'm using Ubuntu 12.04 + Heartbeat 3.0.5-3ubuntu2 to provide high availability for some IP addresses.
> I want to configure an IPv6 address on my haresources. I did this:
> 
> File /etc/heartbeat/haresources:
> 
> server.domain.com \
>    192.168.2.62/32/eth1 \
>    192.168.2.64/32/eth1 \
>    192.168.2.72/32/eth1 \
>    IPv6addr::2001:db8:38a5:8::2006/48/eth1 \
>    MailTo::adm <at> domain.com
> 
> The IPv4 addresses work fine, but I'm not getting success with the IPv6 address.
> My logs shows this message:
> ResourceManager[22129]: info: Running /etc/ha.d/resource.d/IPv6addr 2001:db8:38a5:8
2006/48/eth1 start
> ResourceManager[22129]: CRIT: Giving up resources due to failure of IPv6addr::2001:db8:38a5:8::2006/48/eth1
> ResourceManager[22129]: info: Running /etc/ha.d/resource.d/IPv6addr 2001:db8:38a5:8
2006/48/eth1 stop
> ResourceManager[22129]: info: Retrying failed stop operation [IPv6addr::2001:db8:38a5:8::2006/48/eth1]
> 
> Apparently there is a conflict between the characters '::' inside
> the IPv6 address and the separator '::' used in the haresources. But
> I would not like have to expand the IPv6 address.
> 
> Does anyone know a way to avoid this conflict?
(Continue reading)

Permalink | Reply | 0 comments |
headers
Parkirat | 15 Jun 2013 16:23
Picon

Pacemaker - Resource dont get started on the standby node.

Hi,

I am new pacemaker. I have been trying to setup Linux HA with heartbeat and
pacemaker in ec2.

I started with setting up Apache Server in HA using heartbeat only and
configuring my resource in haresources file, such that when heartbeats in 1
machines stop, the process gets started in the other machines. The only
problem was when the process itself gets stopped, it is not detected unless
the heartbeat also goes off. To fix this, I had to move to pacemaker +
heartbeat.

Now when I kill the process, it gets started on the same machine, but when I
kill the heartbeat itself, the process should move to the other node which
is not happening. 

*Output from crm configure show:*

node $id="091a7142-4272-402a-aeff-ce1b25123668" prod-hb-nmn-002
node $id="7910c4de-718d-45d7-b4da-24b3b65b9855" prod-hb-nmn-001
primitive apache ocf:heartbeat:apache \
	params configfile="/etc/httpd/conf/httpd.conf" \
	op monitor interval="60s" \
	op start interval="0" timeout="40s" \
	op stop interval="0" timeout="60s" \
	meta target-role="Started"
property $id="cib-bootstrap-options" \
	dc-version="1.0.9-89bd754939df5150de7cd76835f98fe90851b677" \
	cluster-infrastructure="Heartbeat" \
	no-quorum-policy="ignore" \
(Continue reading)

Permalink | Reply | 6 comments |
headers
listas | 14 Jun 2013 20:29
Picon

Heartbeat haresources with IPv6

Hi,

I'm using Ubuntu 12.04 + Heartbeat 3.0.5-3ubuntu2 to provide high availability for some IP addresses.
I want to configure an IPv6 address on my haresources. I did this:

File /etc/heartbeat/haresources:

server.domain.com \
    192.168.2.62/32/eth1 \
    192.168.2.64/32/eth1 \
    192.168.2.72/32/eth1 \
    IPv6addr::2001:db8:38a5:8::2006/48/eth1 \
    MailTo::adm <at> domain.com

The IPv4 addresses work fine, but I'm not getting success with the IPv6 address.
My logs shows this message:
ResourceManager[22129]: info: Running /etc/ha.d/resource.d/IPv6addr 2001:db8:38a5:8
2006/48/eth1 start
ResourceManager[22129]: CRIT: Giving up resources due to failure of IPv6addr::2001:db8:38a5:8::2006/48/eth1
ResourceManager[22129]: info: Running /etc/ha.d/resource.d/IPv6addr 2001:db8:38a5:8
2006/48/eth1 stop
ResourceManager[22129]: info: Retrying failed stop operation [IPv6addr::2001:db8:38a5:8::2006/48/eth1]

Apparently there is a conflict between the characters '::' inside the IPv6 address and the separator '::'
used in the haresources. But I would not like have to expand the IPv6 address. 

Does anyone know a way to avoid this conflict?

Thanks!
--
(Continue reading)

Permalink | Reply | 0 comments |
headers
David Vossel | 13 Jun 2013 23:59
Picon
Favicon

ocf HA_RSCTMP directory location

Hey,

Andrew and I have been running into some inconsistencies between resource-agent packages that we need to
get cleared up.

There's an ocf variable, HA_RSCTMP, used in many of the resource agents that represents a place the agent's
can store their PID files and other temporary data.  This data needs to live under some directory in
/var/run as that directory is typically cleared on startup.  This is important to prevent stale PID files
and other transient data from being persistent across restarts.

Anyway.  Here's the problem.

Pacemaker thinks that data should live in '/var/run/heartbeat/rsctmp', but not all the resource-agent
packages are consistent with that.  For example, Suse's resource-agent package sets HA_RSCTMP to
'/var/run/resource-agents' ( looking at this rpm,
http://download.opensuse.org/distribution/11.4/repo/oss/suse/x86_64/resource-agents-1.0.3-9.12.1.x86_64.rpm )

We need to come to some sort of agreement because ultimately Pacemaker needs to make sure this directory
exists on startup, whatever it is.  If pacemaker doesn't create the right directory, it's possible the
resource agents won't be able to access it since /var/run is re-initialized on startup.

so,
HA_RSCTMP = /var/run/heartbeat/rsctmp
or
HA_RSCTMP = /var/run/resource-agents

thoughts?

-- Vossel
(Continue reading)

Permalink | Reply | 5 comments |
headers
Guglielmo Abbruzzese | 13 Jun 2013 19:57
Picon

Pacemaker & gfs2 on RHEL6.2 x64

Hi,
I'm working on a issue from a while but it seems to me I've got to a dead
end :(

Till now, I've been able to setup and configure several cluster solutions on
RHEL6.2x64 using the RH's Corosync (1.4.1-4) and Pacemaker (1.1.6-3) rpms.
I've started to work on installing a SAN storage, I'd like to use a shared
file system because 8 servers need to store something on it at the same
time.
OCFS2 seems to be not available anymore for my distro. NFS is not suitable.
So I decided to try and use gfs2, even because of the excellent
"Pacemaker-1.1-Clusters_from_Scratch" install and configuration guide.

Here are some troubles.

1. The user guide refers to Fedora13. It suggests to install  gfs2-utils,
cman and gfs2-cluster. In particular, gfs2-cluster is not provided with
RHEL, and the version I've found  (gfs2-cluster-3.1.1-2.fc14.x86_64.rpm)
requires cman >= 3.1.0-1 and  openais >= 1.1.2-1; not even RHEL6.4 has got
such versions. What's the best practice in this case?

2. Pacemaker is provided with RHEL and so I thought it was supported in some
way by the RHEL's team (at least about the compatibility with the other
packages); but it isn't actually, and the support has strongly suggested to
swap to Red Hat Cluster Suite. I've been working a lot on Pacemaker, this is
a choice I really wouldn't take...

3. In the worst case, any suggestion about mounting and managing a shared
SAN storage file system  with a different solution without trashing all my
job till now for switching to RHCS?
(Continue reading)

Permalink | Reply | 1 comment |
headers
Ulrich Windl | 13 Jun 2013 13:57
Picon

"Use of uninitialized value in numeric gt (>) at /usr/sbin/ldirectord line 4037."

Hi!

In SLES11 SP2 ldirectord (ldirectord-3.9.4-0.26.84) in debug mode shows the message:
Use of uninitialized value in numeric gt (>) at /usr/sbin/ldirectord line 4037.

It looks like a bug to me. The line in question (in _service_up()) is this:
 if ($r->{failcount} > 0) {

Only AFTER that line, that variable seems to be set:
$r->{failcount} = 0;

It does not look like a serious problem, but the message is clearly avoidable by initializing the variable
before using it.

(personally I develop all my Perl with "use warnings; use strict")

Regards,
Ulrich

_______________________________________________
Linux-HA mailing list
Linux-HA <at> lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
Permalink | Reply | 1 comment |

Gmane