Mailing List Archive: Linux-HA: Users

groups and colocation (lsb script and ip)

Index | Next | Previous | View Threaded

jeff at bofus

Dec 6, 2007, 7:55 PM

Post #1 of 5 (2560 views)
Permalink

groups and colocation (lsb script and ip)

I have created a simple heartbeat cluster:

2 Centos 4.5 nodes

HB version:
heartbeat-pils-2.1.2-3.el4.centos
heartbeat-stonith-2.1.2-3.el4.centos
heartbeat-gui-2.1.2-3.el4.centos
heartbeat-2.1.2-3.el4.centos

Here is the issue I see, and I dont know what I am doing wrong.
I have a simple group (*"GROUP_knworks_assp"*) setup. It is colocated
and ordered.
the issue is that I can get the IP address to work, but the LSB init
script will not work.

I have followed the document:
http://www.linux-ha.org/LSBResourceAgent
it meets all of the requirements and passes all the correct return codes
to be LSB Compliant.

The IP address is on node1, and the LSB script says: unmanaged and failed.

crm_mon -1 output:
Resource Group: GROUP_knworks_assp
knworks_assp_ip (heartbeat::ocf:IPaddr2): Started asknmapr01
knworks_assp_initscript (lsb:knworks_assp_init_script.sh):
Started asknmapr02 (unmanaged) FAILED

Failed actions:
knworks_assp_initscript_monitor_0 (node=asknmapr02, call=77,
rc=254): Error
knworks_assp_initscript_stop_0 (node=asknmapr02, call=78, rc=254): Error

Here is the resources section of my CIB:
<resources>
<primitive class="ocf" type="IPaddr2" provider="heartbeat"
id="IPADDR_WEB">
<instance_attributes id="IPADDR_WEB_instance_attrs">
<attributes>
<nvpair id="e4884172-6b9d-4d60-88d2-99ff34756ac5" name="ip"
value="10.1.2.190"/>
<nvpair name="target_role" id="IPADDR_WEB_target_role"
value="started"/>
</attributes>
</instance_attributes>
</primitive>
* <group id="GROUP_knworks_assp" ordered="true" collocated="true">
<primitive id="knworks_assp_ip" class="ocf" type="IPaddr2"
provider="heartbeat">
<instance_attributes id="knworks_assp_ip_instance_attrs">
<attributes>
<nvpair id="cfffb54c-3b2c-4e32-9ded-b78142aa4940" name="ip"
value="10.1.2.235"/>
</attributes>
</instance_attributes>
</primitive>
<primitive id="knworks_assp_initscript" class="lsb"
type="knworks_assp_init_script.sh" provider="heartbeat">
<instance_attributes id="knworks_assp_initscript_instance_attrs">
<attributes>
<nvpair id="knworks_assp_initscript_target_role"
name="target_role" value="started"/>
</attributes>
</instance_attributes>
</primitive>
</group>*
</resources>

Any help would be appreciated and save my forehead, I have been banging
it on the table for 2 days.

Thanks!

-Jeff Humes
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

beekhof at gmail

Dec 7, 2007, 2:34 AM

Post #2 of 5 (2431 views)
Permalink

Re: groups and colocation (lsb script and ip) [In reply to]

On Dec 7, 2007, at 4:55 AM, Jeff Humes wrote:

> I have created a simple heartbeat cluster:
>
> 2 Centos 4.5 nodes
>
> HB version:
> heartbeat-pils-2.1.2-3.el4.centos
> heartbeat-stonith-2.1.2-3.el4.centos
> heartbeat-gui-2.1.2-3.el4.centos
> heartbeat-2.1.2-3.el4.centos
>
>
> Here is the issue I see, and I dont know what I am doing wrong.
> I have a simple group (*"GROUP_knworks_assp"*) setup. It is
> colocated and ordered.
> the issue is that I can get the IP address to work, but the LSB init
> script will not work.
>
> I have followed the document:
> http://www.linux-ha.org/LSBResourceAgent
> it meets all of the requirements and passes all the correct return
> codes to be LSB Compliant.
>
> The IP address is on node1, and the LSB script says: unmanaged and
> failed.
>
> crm_mon -1 output:
> Resource Group: GROUP_knworks_assp
> knworks_assp_ip (heartbeat::ocf:IPaddr2): Started
> asknmapr01
> knworks_assp_initscript
> (lsb:knworks_assp_init_script.sh): Started asknmapr02
> (unmanaged) FAILED
>
> Failed actions:
> knworks_assp_initscript_monitor_0 (node=asknmapr02, call=77,
> rc=254): Error
> knworks_assp_initscript_stop_0 (node=asknmapr02, call=78, rc=254):
> Error
>

254? that can't be good - especially since thats supposed to be the
massaged rc value (the raw rc is mapped to one of the entries in the
UNIFORM_RET_EXECRA enum).

what do the logs say?

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

jeff at bofus

Dec 7, 2007, 10:27 AM

Post #3 of 5 (2426 views)
Permalink

Re: groups and colocation (lsb script and ip) [In reply to]

I figure out this issue just after I compile this email.
the second node did not have the link in /etc/init.d for the lsb script.

All seems to be working now.

I am confused a little about this this works though. When I start a
resource on one node, does it automatically run the stop of that
resource on the other nodes?

-JPH

Andrew Beekhof wrote:
>
> On Dec 7, 2007, at 4:55 AM, Jeff Humes wrote:
>
>> I have created a simple heartbeat cluster:
>>
>> 2 Centos 4.5 nodes
>>
>> HB version:
>> heartbeat-pils-2.1.2-3.el4.centos
>> heartbeat-stonith-2.1.2-3.el4.centos
>> heartbeat-gui-2.1.2-3.el4.centos
>> heartbeat-2.1.2-3.el4.centos
>>
>>
>> Here is the issue I see, and I dont know what I am doing wrong.
>> I have a simple group (*"GROUP_knworks_assp"*) setup. It is
>> colocated and ordered.
>> the issue is that I can get the IP address to work, but the LSB init
>> script will not work.
>>
>> I have followed the document:
>> http://www.linux-ha.org/LSBResourceAgent
>> it meets all of the requirements and passes all the correct return
>> codes to be LSB Compliant.
>>
>> The IP address is on node1, and the LSB script says: unmanaged and
>> failed.
>>
>> crm_mon -1 output:
>> Resource Group: GROUP_knworks_assp
>> knworks_assp_ip (heartbeat::ocf:IPaddr2): Started asknmapr01
>> knworks_assp_initscript (lsb:knworks_assp_init_script.sh):
>> Started asknmapr02 (unmanaged) FAILED
>>
>> Failed actions:
>> knworks_assp_initscript_monitor_0 (node=asknmapr02, call=77,
>> rc=254): Error
>> knworks_assp_initscript_stop_0 (node=asknmapr02, call=78, rc=254):
>> Error
>>
>
> 254? that can't be good - especially since thats supposed to be the
> massaged rc value (the raw rc is mapped to one of the entries in the
> UNIFORM_RET_EXECRA enum).
>
> what do the logs say?
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

beekhof at gmail

Dec 8, 2007, 2:45 AM

Post #4 of 5 (2426 views)
Permalink

Re: groups and colocation (lsb script and ip) [In reply to]

On Dec 7, 2007 7:27 PM, Jeff Humes <jeff [at] bofus> wrote:
> I figure out this issue just after I compile this email.
> the second node did not have the link in /etc/init.d for the lsb script.
>
> All seems to be working now.
>
> I am confused a little about this this works though. When I start a
> resource on one node, does it automatically run the stop of that
> resource on the other nodes?

Not quite.

We check to make sure its not already running before we start it anywhere.
In your case, this check would have failed (since the script didn't
exist) and then the stop would have failed... etc etc

>
> -JPH
>
>
> Andrew Beekhof wrote:
> >
> > On Dec 7, 2007, at 4:55 AM, Jeff Humes wrote:
> >
> >> I have created a simple heartbeat cluster:
> >>
> >> 2 Centos 4.5 nodes
> >>
> >> HB version:
> >> heartbeat-pils-2.1.2-3.el4.centos
> >> heartbeat-stonith-2.1.2-3.el4.centos
> >> heartbeat-gui-2.1.2-3.el4.centos
> >> heartbeat-2.1.2-3.el4.centos
> >>
> >>
> >> Here is the issue I see, and I dont know what I am doing wrong.
> >> I have a simple group (*"GROUP_knworks_assp"*) setup. It is
> >> colocated and ordered.
> >> the issue is that I can get the IP address to work, but the LSB init
> >> script will not work.
> >>
> >> I have followed the document:
> >> http://www.linux-ha.org/LSBResourceAgent
> >> it meets all of the requirements and passes all the correct return
> >> codes to be LSB Compliant.
> >>
> >> The IP address is on node1, and the LSB script says: unmanaged and
> >> failed.
> >>
> >> crm_mon -1 output:
> >> Resource Group: GROUP_knworks_assp
> >> knworks_assp_ip (heartbeat::ocf:IPaddr2): Started asknmapr01
> >> knworks_assp_initscript (lsb:knworks_assp_init_script.sh):
> >> Started asknmapr02 (unmanaged) FAILED
> >>
> >> Failed actions:
> >> knworks_assp_initscript_monitor_0 (node=asknmapr02, call=77,
> >> rc=254): Error
> >> knworks_assp_initscript_stop_0 (node=asknmapr02, call=78, rc=254):
> >> Error
> >>
> >
> > 254? that can't be good - especially since thats supposed to be the
> > massaged rc value (the raw rc is mapped to one of the entries in the
> > UNIFORM_RET_EXECRA enum).
> >
> > what do the logs say?
> >
>
> > _______________________________________________
> > Linux-HA mailing list
> > Linux-HA [at] lists
> > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

jeff at bofus

Dec 8, 2007, 6:36 AM

Post #5 of 5 (2418 views)
Permalink

Re: groups and colocation (lsb script and ip) [In reply to]

I appreciate the clarification.
Thanks to you all for a great product!

-JPH

Andrew Beekhof wrote:
> On Dec 7, 2007 7:27 PM, Jeff Humes <jeff [at] bofus> wrote:
>
>> I figure out this issue just after I compile this email.
>> the second node did not have the link in /etc/init.d for the lsb script.
>>
>> All seems to be working now.
>>
>> I am confused a little about this this works though. When I start a
>> resource on one node, does it automatically run the stop of that
>> resource on the other nodes?
>>
>
> Not quite.
>
> We check to make sure its not already running before we start it anywhere.
> In your case, this check would have failed (since the script didn't
> exist) and then the stop would have failed... etc etc
>
>
>> -JPH
>>
>>
>> Andrew Beekhof wrote:
>>
>>> On Dec 7, 2007, at 4:55 AM, Jeff Humes wrote:
>>>
>>>
>>>> I have created a simple heartbeat cluster:
>>>>
>>>> 2 Centos 4.5 nodes
>>>>
>>>> HB version:
>>>> heartbeat-pils-2.1.2-3.el4.centos
>>>> heartbeat-stonith-2.1.2-3.el4.centos
>>>> heartbeat-gui-2.1.2-3.el4.centos
>>>> heartbeat-2.1.2-3.el4.centos
>>>>
>>>>
>>>> Here is the issue I see, and I dont know what I am doing wrong.
>>>> I have a simple group (*"GROUP_knworks_assp"*) setup. It is
>>>> colocated and ordered.
>>>> the issue is that I can get the IP address to work, but the LSB init
>>>> script will not work.
>>>>
>>>> I have followed the document:
>>>> http://www.linux-ha.org/LSBResourceAgent
>>>> it meets all of the requirements and passes all the correct return
>>>> codes to be LSB Compliant.
>>>>
>>>> The IP address is on node1, and the LSB script says: unmanaged and
>>>> failed.
>>>>
>>>> crm_mon -1 output:
>>>> Resource Group: GROUP_knworks_assp
>>>> knworks_assp_ip (heartbeat::ocf:IPaddr2): Started asknmapr01
>>>> knworks_assp_initscript (lsb:knworks_assp_init_script.sh):
>>>> Started asknmapr02 (unmanaged) FAILED
>>>>
>>>> Failed actions:
>>>> knworks_assp_initscript_monitor_0 (node=asknmapr02, call=77,
>>>> rc=254): Error
>>>> knworks_assp_initscript_stop_0 (node=asknmapr02, call=78, rc=254):
>>>> Error
>>>>
>>>>
>>> 254? that can't be good - especially since thats supposed to be the
>>> massaged rc value (the raw rc is mapped to one of the entries in the
>>> UNIFORM_RET_EXECRA enum).
>>>
>>> what do the logs say?
>>>
>>>
>>> _______________________________________________
>>> Linux-HA mailing list
>>> Linux-HA [at] lists
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>>>
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
>>
>>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Index | Next | Previous | View Threaded

Interested in having your list archived? Contact Gossamer Threads