Hi,
This seems the same problem I reported a while ago. It was fixed in
http://hg.clusterlabs.org/pacemaker/stable-1.0/rev/214f0fc258f2.
Thanks
On Fri, Mar 12, 2010 at 2:36 PM, Quentin Smith wrote:
> Hi-
>
> I just took the latest updates to pacemaker and heartbeat from
> http://people.debian.
On Tue, Mar 9, 2010 at 12:20 AM, Andrew Beekhof wrote:
> On Mon, Mar 8, 2010 at 9:37 PM, hj lee wrote:
> > Hi,
> >
> > In the typical multi-state resource agent, it changes master score on
> demote
> > or promote. Each change in master score triggers PE calculatio
I don't think it's a good idea to put location on clone. The clone is
designed to run equally on every node. Please remove these and see that
helps.
location LDAP-IP-placement-2 LDAP-IP 50: genome-ldap2
location LDAP-placement-1 LDAP-clone 100: genome-ldap1
location LDAP-placement-2 LDAP-clone 100
Hi,
In the typical multi-state resource agent, it changes master score on demote
or promote. Each change in master score triggers PE calculation. Suppose the
following scenario.
1. Pacemaker initiates demote/promote
2. demote is issued and lower the master score on the demoted node.
3. promote is
On Fri, Jan 29, 2010 at 1:46 AM, Andrew Beekhof wrote:
> On Tue, Jan 19, 2010 at 5:57 PM, hj lee wrote:
> >
> >
> > On Tue, Jan 19, 2010 at 9:08 AM, Andrew Beekhof
> wrote:
> >>
> >> I ran your script, all I get is:
> >>
> >> [r...@pc
On Tue, Jan 19, 2010 at 9:08 AM, Andrew Beekhof wrote:
> I ran your script, all I get is:
>
> [r...@pcmk-4 ~]# ./test.sh 0
> scope=status name=master-vmrd-res:0 value=(null)
> Error performing operation: The object/attribute does not exist
> scope=status name=master-vmrd-res:0 value=1
> scope=s
On Tue, Jan 12, 2010 at 2:15 PM, hj lee wrote:
>
>
>> 2010-01-12 00:19:47.879800 copper1-cib: [4764]: debug: cib_process_xpath:
>> Processing cib_query op for //cib/status//node_sta...@id='
>> copper1.dr.peach.com']//nvpa...@name='master-vm
On Tue, Jan 12, 2010 at 1:37 PM, hj lee wrote:
>
>
> It happened again, I attched cib files and log messages of both nodes. The
> problem happened last night, but cib file was captured this morning. As you
> can see below log expert, the attrd receives 2 entries for cib_query,
On Mon, Jan 11, 2010 at 2:45 AM, Andrew Beekhof wrote:
>
> >>
> >> When I checked the cib, it has only one entry for each host.
> >>
> >> [r...@silverthorne2 tmp]# cibadmin -Ql | grep master-vmrd
> >>>> id="status-copper2-master-vmrd-res:0" value="5"/>
> >>>> id="status-co
On Mon, Jan 11, 2010 at 1:22 PM, hj lee wrote:
> Hi,
>
> I have a situation crm_master does not update a master score. I increased
> log level at attrd. And I found that current value is 2 but stored value
> is 5. I don't know how attrd ends up this state. But attrd ignor
Hi,
I have a situation crm_master does not update a master score. I increased
log level at attrd. And I found that current value is 2 but stored value
is 5. I don't know how attrd ends up this state. But attrd ignores this
update, so a master score does not change at all. I think the code shou
On Thu, Jan 7, 2010 at 12:49 PM, hj lee wrote:
>
>
> On Thu, Jan 7, 2010 at 5:24 AM, Andrew Beekhof wrote:
>
>> On Wed, Jan 6, 2010 at 3:09 AM, hj lee wrote:
>> > Hi,
>> >
>> > I have a very simple master/slave in two node cluster. The crm_
On Thu, Jan 7, 2010 at 5:24 AM, Andrew Beekhof wrote:
> On Wed, Jan 6, 2010 at 3:09 AM, hj lee wrote:
> > Hi,
> >
> > I have a very simple master/slave in two node cluster. The crm_attribute
> > failed to master score because find_nvpair_attr() found multiple matches.
Hi,
I have a very simple master/slave in two node cluster. The crm_attribute
failed to master score because find_nvpair_attr() found multiple matches.
But I think that's not possible, there should be always one entry for master
score for each node. I just had one failure today. The log is below. I
On Sun, Jan 3, 2010 at 11:14 AM, Martin Gombač wrote:
> Hi,
>
> i removed colocation constraint per your suggestion.
> Please find other requested information attached.
> As logs show ibm1 restarted resource Hosting, when ibm2 was shut down,
> while leaving underlying drbd resource promoted. I wi
> - Have you installed openais package?
>>> - ocfs2_controld is using the CKPT service which is provided by openais
>>> - not corosync.
>>>
>>> really? I did not know that. So my best bet is either switch to openais
>>> or
>>>
>>
>> You might misunderstand what I said. It is not "switch to openais"
On Thu, Dec 31, 2009 at 3:06 AM, Martin Gombač wrote:
>
>> First why do you set max-master=2? So both nodes are promoted and both
>> nodes are master. Then your colocation constraints does not make sense.
>> Because Hosting has to be started on both nodes. If this is not what you
>> want, then pl
I found a few problems in your configuration.
First why do you set max-master=2? So both nodes are promoted and both nodes
are master. Then your colocation constraints does not make sense. Because
Hosting has to be started on both nodes. If this is not what you want, then
please remove max-master=
On Tue, Dec 29, 2009 at 11:09 AM, Daniel Qian wrote:
>
>> On Mon, Dec 28, 2009 at 07:05:24PM -0500, Daniel Qian wrote:
>>
>>> I am using Corosync 1.2.0 that comes with Fedora 12 and have this
>>> error when trying to start corosync service. If I set it to anything
>>> above 1200 the error goes awa
Hi,
Maybe this is related to no-quorum-policy. What is no-quorum-policy? You can
check no-quorum-policy in "crm configure show" command. If you can not see
in "crm configure show" command, then it is "stop" by default. If that is
your case, then please set it to ignore by "crm configure property
n
On Wed, Dec 23, 2009 at 8:34 AM, Dejan Muhamedagic wrote:
>
> > > > 3. Why the pingd attribute was not set immediately after pingd
> started
> > > > up, and was able to ping the ping node. After the pingd was started,
> then
> > > > it waited 60 seconds (the timeout value) to set the attribute so
On Wed, Dec 16, 2009 at 12:52 PM, Frank DiMeo wrote:
> Hi All,
>
> Does anyone know the correct syntax for the crm_master command when used
> from a shell? I want to promote (and at some point demote) one of my
> multistate resources running on one of my nodes, but haven't found the
> correct syn
On Thu, Dec 17, 2009 at 8:23 AM, Frank DiMeo wrote:
> Hi,
>
>
>
> I need a way to inform the cluster of failures between monitor polls of my
> multistate resources, and I notice that I cannot use “crm_resource “ to
> inform the cluster of the failure of a master (multistate) resource, as it
> see
Hi,
I have two-node cluster with a simple multi-state clone. I am using "crm_mon
-1 | grep Masters: | awk '{print $3}' " to find out which node is a master.
Is there a better way doing this? I don't like relying on the output of
crm_mon command, because the output can change in the future.
Thanks
On Wed, Dec 16, 2009 at 3:53 AM, Andrew Beekhof wrote:
> On Wed, Dec 16, 2009 at 2:24 AM, Javen Wu wrote:
>
> As per the conversation on IRC, DC election has nothing to do with quorum.
> But only the DC that also has quorum is able to fence the unseen nodes
> and run resources.
>
> In two-node c
Hi,
I want to use logical AND of two expresson in rule. For example, I want to
create location constraint like this. Assume there is pingd and hostname.
crm configure location my-test-location my-test-resource rule inf: pingd
defined AND #uname eq hostname1
How can I do this?
Thanks
hj
Hi,
I want to monitor some of resources less than one sec interval. I tried to
set "100ms" or "0.1s", it does not work.
Thanks
hj
___
Pacemaker mailing list
Pacemaker@oss.clusterlabs.org
http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> I wanted to test the split-brain situation too and the recovery from it. I
> also wanted to test a pingd resource and location we also have configured to
> see it the node put down the resources correctly when it detects no
> connection to the gateway.
>
> Anyway, I have checked this situation an
On Fri, Nov 27, 2009 at 3:05 PM, Steven Dake wrote:
> On Fri, 2009-11-27 at 11:32 -0200, Mark Horton wrote:
> > I'm using pacemaker 1.0.6 and corosync 1.1.2 (not using openais) with
> > centos 5.4. The packages are from here:
> > http://www.clusterlabs.org/rpm/epel-5/
> >
> > Mark
> >
> > On Fri
On Fri, Nov 20, 2009 at 12:42 AM, Andrew Beekhof wrote:
> >
> > I need a resource to do something on event of node join/left. Is it
> possible
> > to receive node join/left event at resource level?
>
> If its a clone, then you can ask for notifications.
> Otherwise, you need to look in the cib or
Hi,
I need a resource to do something on event of node join/left. Is it possible
to receive node join/left event at resource level?
Thanks
hj
___
Pacemaker mailing list
Pacemaker@oss.clusterlabs.org
http://oss.clusterlabs.org/mailman/listinfo/pacemaker
AM +0100, Andrew Beekhof wrote:
> > >> On Wed, Nov 18, 2009 at 11:04 PM, hj lee wrote:
> > >> > Hi,
> > >> >
> > >> > Sorry for long delay. This is my pingd configuration.
> > >> >
> > >> > primitive pingd-res o
Hi again,
I found what is wrong. I should use "monitor" as an operation name in crm
command. I though it is id, realized today it becomes operation name
directly.
Thanks
hj
On Wed, Nov 18, 2009 at 3:04 PM, hj lee wrote:
> Hi,
>
> Sorry for long delay. This is my
ION" | grep monitor > /dev/null; then
__OCF_ACTION=monitor
fi
Thanks
On Mon, Nov 16, 2009 at 7:34 AM, Andrew Beekhof wrote:
> On Thu, Nov 12, 2009 at 9:02 PM, hj lee wrote:
> >
> >
> > On Thu, Nov 12, 2009 at 3:58 AM, Lars Marowsky-Bree wrote:
> >>
&
On Wed, Nov 18, 2009 at 2:27 PM, Frank DiMeo wrote:
> Thanks for you response, some comments:
>
>
>
> 1) So, if we cannot load a persistent configuration after a restart
> of the all nodes in the cluster, how does one automatically configure an
> (rebooted) unattended cluster with a known
> 1) Is there a way to start pacemaker and specify an xml file as the startup
> configuration? Right now, I start openais/pacemaker, wait for DC election,
> then run cibadmin to load the file. That's fine for now, but I want to
> automate this eventually.
>
I don't think it's possible.
> 2) Is
t; simulations and public IPs for the real servers, but I don't think it is
> important.
>
> Thanks for your patience,
> Giovanni
>
>
>
> On Nov 13, 2009, at 1:36 PM, hj lee wrote:
>
> Hi,
>
> I have the same problem in CentOS 5.3 with pacemaker-1.0.5 and
>
Hi,
I have the same problem in CentOS 5.3 with pacemaker-1.0.5 and
openais-0.80.5. This is openais bug! Two problems.
1. Starting openais service gets seg fault sometime. It more likely happens
if openais service get started before syslog.
2. The seg fault handler of openais calls syslog(). The sy
On Thu, Nov 12, 2009 at 3:58 AM, Lars Marowsky-Bree wrote:
> On 2009-11-11T16:48:36, hj lee wrote:
>
> > Here rewriting again with correct name:
> >
> > I am using pacemaker-1.0.5 on CentOS 5.3 with openais cluster stack. I
> set
> > up ocf:hearbeat:pin
One correction: The variable name $__ACTION in my previous mail was wrong,
the correct name is $__OCF_ACTION.
Here rewriting again with correct name:
I am using pacemaker-1.0.5 on CentOS 5.3 with openais cluster stack. I set
up ocf:hearbeat:pingd as a clone and enabled monitor operation. The moni
Hi,
I am using pacemaker-1.0.5 on CentOS 5.3 with openais cluster stack. I set
up ocf:hearbeat:pingd as a clone and enabled monitor operation. The monitor
operation returns $OCF_ERR_UNIMPLEMENTED. I printed $__ACTION variable in
pingd RA, surprisingly that variable is set to pingd-monitor-op, not
Hi,
When I stop a multi-state clone resource, first the master becomes slave,
and then it is stopped. This happens also when I do "service openais stop".
But I want the master gets stopped without demoting to slave. Is there any
specific reason for current design? Is there any way to configure to
Hi,
I configured two nodes cluster on RHEL 5.3 with the following resources.
Note that I am using pacemaker-1.0.6.
- IPMI stonith as a clone. Each IPMI clone is monitoring the other node.
- One Master/Slave resource: Master is running on node1, Slave is running on
node2.
- One FakeIPMI resource.
Thank you, it works OK now!
On Mon, Oct 12, 2009 at 1:31 AM, Andrew Beekhof wrote:
> On Fri, Oct 9, 2009 at 10:08 PM, hj lee wrote:
> > Hi Andrew,
> >
> > According to the Pacemaker Configuration Explained, the meta attribute
> and
> > instance attribute object c
Hi Andrew,
According to the Pacemaker Configuration Explained, the meta attribute and
instance attribute object can have rule inside of it. But cibadmin command
does not accept it. Is it a documentation error or bug in software?
Thank you
On Thu, Oct 8, 2009 at 6:11 PM, hj lee wrote:
>
does not accept rule same as meta_attributes.
Is there any other way to set a different value to instance attribute?
Thanks
On Thu, Oct 8, 2009 at 2:47 PM, hj lee wrote:
> Hi,
>
> The XML below is an example in Configuration Explained 1.0 section of
> "Using Rules to Control Res
Hi,
The XML below is an example in Configuration Explained 1.0 section of "Using
Rules to Control Resource Options. Is there any way to create the same
instance_attribute by crm command?
Thanks
.
___
Pacemaker mailing list
Pacemaker@oss.clus
Yes, it's my syslog configuration. After adding daemon.* to syslog.conf, it
shows up now.
Thanks very much
On Mon, Aug 24, 2009 at 12:19 PM, Andrew Beekhof wrote:
> could depend on your syslog configuration... it works here
>
> On Mon, Aug 24, 2009 at 7:27 PM, hj lee wrote:
> &
Hi,
I am trying to increase the log level in pengine. I sent may USR1 signal to
pengine process, but still debug level logs are never printed in the syslog.
#ps -ax
24356 pts/15 S 0:00 /usr/lib/heartbeat/attrd
24357 pts/15 S 0:00 /usr/lib/heartbeat/pengine
24358 pts/15 S 0:01
Hi,
After upgrading pacemaker 1.0.5, ordering_type2text reports many errors. But
these types are valid types defined in pengine.h. The checking function does
not check all the types defined.
Thank you
Aug 24 09:21:44 silverthorne2 pengine: [24357]: ERROR: ordering_type2text:
Unknown ordering typ
Hi,
I had the same problem after upgrading to pacemaker 1.0.5 in RHLE 5.3. After
deleting all the files in /var/lib/pengine/ directory, this problem seems
gone, I haven't seen it so far. Maybe it is related the UID change in
pengine(hacluster to daemon) in 1.0.5, but not exactly sure.
hj
On Fri,
Is the rebooted node host2? Then it makes sense because location score is
INFINITE in host2.
hj
On Thu, Aug 20, 2009 at 10:14 AM, Gerry kernan
wrote:
> Hi
>
>
>
> I have a 2 node cluster using openais & pacemaker,
>
>
>
> If I reboot the standby node when the node comes back up all resources
>
...@sonet.internetdsl.pl <
cu...@sonet.internetdsl.pl> wrote:
> hj lee pisze:
>
>> What is serwer-mysql? In the last location constraint, serwer-mysql should
>> be msDRBD I think. Does msDRBD come up master and slave? Or both came up
>> with slave?
>>
>> Thanks
&
What is serwer-mysql? In the last location constraint, serwer-mysql should
be msDRBD I think. Does msDRBD come up master and slave? Or both came up
with slave?
Thanks
On Wed, Aug 19, 2009 at 1:25 AM, wrote:
> Hi, i would like to make following configuration: DRBD as master
> on two nodes +OCFS2
e node goes down, there is no quorum, so the
> resources will not be migrated. You need to have this entry so that one node
> can operate on its own. This was happening to my 2 node cluster, I would
> shutdown one node and the second one would not get promoted to master, that
> was until I
Thank very much for the reply.
I tested it both stonith-enabled and no-quorum-policy. As Dejan pointed,
this is related to stonith-enabled. With stonith-enabled true (which is
default),
if I kill the master node, the slave stays as a slave, it seems expecting
something from stonith. With stonith-e
One more error I found is in the table of Available Cluster Options on page
9.
This table says the default value of stonith-enabled is FALSE. But if I
start heartbeat or openais with empty cib.xml, the crm_gui shows
"stonith-enabled" checked (means enabled) in the "Policy Engine" tab of "CRM
Confi
Hi,
I defined stonith:ssh, they are running in both machines as a clone. How is
the stonith related to promoting standby? When the Pacemaker detects master
node was gone in cluster, then why doesn't Pacemaker promote the standby?
Thanks very much
On Mon, Aug 17, 2009 at 4:00 PM, Dejan Muhamedagi
Hi,
I am very simple Master/Slave in RHEL 5.3 with pacemaker 1.0.4 and heartbea
2.99. I unplugged the power cable at the master machine, and I expected the
slave becomes master. But the slave stays at slave state. Is this correct
behavior or a bug? How does Pacemaker (or heartbeat) handle this kin
Hi,
I think you will need order and colocation constraints. Please read DRBD
Howto http://www.clusterlabs.org/wiki/DRBD_HowTo_1.0.
Thanks
On Mon, Aug 10, 2009 at 10:56 AM, Michal wrote:
> Hello
> I have following error while running pacemaker-openais:
>
> Failed actions:
> fs0_monitor_0 (node
The trunk is working OK now. The pcmk initialization code gets nodeid
correctly! Thanks for the fix.
hj
On Mon, Jul 27, 2009 at 10:43 AM, Jonathan wrote:
> Not yet. I reverted to corosync/openais 1.0.0 and the issue went away,
> so I am continuing my configuration and testing with those version
Thank you very much for your reply.
When I use the heartbeat stack, it takes 3 sec. In the same environment, it
takes less than 1 sec with OpenAIS stack. I am using RHEL 5 rpms from
Pacemaker and Pacemaker 1.0.4. Did any body measure the time takes from
demote to promote on both stacks? I wonder w
Hi,
I am building a simple multi-state clone with max-master=2. The
demote/promote is working OK by using migration-threshold or co-location
constraint. But I have two questions:
1. Except these two methods, is there any way to invoke demote/promote
manually just for testing?
2. By using co-locat
Hi,
I think the drbd RA can not have two masters! So master-max should be one.
Thanks
hj
On Fri, Jul 17, 2009 at 12:49 PM, Michael Schwartzkopff
wrote:
> Hi,
>
> I configured a multistate DRBD (dual master) and OCFS2 on my cluster. I
> want to
> mount the filesystem where DRBD is master. The fo
64 matches
Mail list logo
