[Pacemaker] problem configuring DRBD resource in "Floating Peers" mode

Wed May 27 23:36:31 EDT 2009

Hi again!
 I forgot to share my changed configurations. Please see attachments.
Thank you for the support!

From: Димитър Бойн 

Sent: Wednesday, May 27, 2009 8:26 PM

 To: 'pacemaker at clusterlabs.org'

 Subject: RE: [Pacemaker] problem configuring DRBD resource in  Floating Peers  mode

Thank you Raoul, for sharing these important! Changes!

Now my #crm_mon -r -V -i 2 looks better:
"
============
Last updated: Wed May 27 20:14:45 2009
Current DC: c001mlb_node01b - partition with quorum
Version: 1.0.3-b133b3f19797c00f9189f4b66b513963f9d25db9
26 Nodes configured, 2 expected votes
3 Resources configured.
============

Online: [ c001mlb_node01a c001mlb_node01b ]
OFFLINE: [ c001mlb_node02a c001mlb_node03a c001mlb_node04a c001mlb_node05a c001mlb_node06a c001mlb_node07a c001mlb_node08a c001mlb_node09a c001mlb_node10a c
001mlb_node11a c001mlb_node12a c001mlb_node13a c001mlb_node02b c001mlb_node03b c001mlb_node04b c001mlb_node05b c001mlb_node06b c001mlb_node07b c001mlb_node0
8b c001mlb_node09b c001mlb_node10b c001mlb_node11b c001mlb_node12b c001mlb_node13b ]

Full list of resources:

ip-c001drbd01a  (ocf::heartbeat:IPaddr2):       Started c001mlb_node01a
ip-c001drbd01b  (ocf::heartbeat:IPaddr2):       Started c001mlb_node01b
Master/Slave Set: ms-drbd0
        Stopped: [ drbd0:0 drbd0:1 ]

Failed actions:
    drbd0:0_start_0 (node=c001mlb_node01a, call=6, rc=1, status=complete): unknown error
    drbd0:0_start_0 (node=c001mlb_node01b, call=6, rc=1, status=complete): unknown error
"

I tried changing "clone-overrides-hostname ="no"and setting drbd.conf to the correct "uname -n"values and it was working!

Has anybody succeed recently to configure floating peers?
Also Andrew Beekhof [andrew at beekhof.net] told me last night that "Its no longer possible to create colocation constraints for a specific 
clone instance" -Which makes impossible to enforce limitations like:
<constraints>
<rsc_colocation id= colocate-drbd0-ip-c001drbd01a  rsc= ip-c001drbd01a  with-rsc= ms-drbd0:0  score= INFINITY />
<rsc_colocation id= colocate-drbd0-ip-c001drbd01b  rsc= ip-c001drbd01b  with-rsc= ms-drbd0:1  score= INFINITY />
</constraints>

To my view it becomes impossible to assure correct resource collocations between two sites in the context of DRBD -am I missing something?

From: Raoul Bhatia [IPAX] [mailto:r.bhatia at ipax.at] 

Sent: Wednesday, May 27, 2009 1:33 AM

 To: pacemaker at clusterlabs.org

 Cc: pacemaker at clusterlabs.org

 Subject: Re: [Pacemaker] problem configuring DRBD resource in  Floating Peers  mode

dДимитър Бойн wrote:
>Hi,
>
>My ultimate goal is to run a bunch of servers/nodes that shall be able 
>to handle a bunch of floating drbd peers.

the last time i tried to use floating peers, it did not work out properly - especially when using outdating/.. mechanisms.

i recall a discussion last year [1] but cannot remember the conclusion.

i saw your colocation request. did you check/resolve that?

i'll have a look anyways ...

>c001mlb_node01a:root >cat /etc/drbd.conf
>
># /usr/share/doc/drbd82/drbd.conf
looks fine.

>c001mlb_node01a:root >cibadmin -Q
*snip*
>
>      <master id= ms-drbd0 >
>
>        <meta_attributes id= ma-ms-drbd0 >
>
>          <nvpair id= ma-ms-drbd0-1 name= clone_max value= 2 />
all  _  have been replaced by  -  -> clone_max is now clone-max. please recheck your configuration to comply with pacemaker-1.0 and above.

imho, that is the reason for having 26 clone instances of each drbd
device: drbd0:0 to drbd0:25.

moreover, and withouth looking further down, i would say that this might fix some of your issues.

>
>          <nvpair id= ma-ms-drbd0-2 name= clone_node_max value= 1 />
>
>          <nvpair id= ma-ms-drbd0-3 name= master_max value= 1 />
>
>          <nvpair id= ma-ms-drbd0-4 name= master_node_max 
>value= 1 />
>
>          <nvpair id= ma-ms-drbd0-5 name= notify value= yes />
>
>          <nvpair id= ma-ms-drbd0-6 name= globally_unique 
>value= true />
>
>          <nvpair id= ma-ms-drbd0-7 name= target_role 
>value= started />
>
>        </meta_attributes>
>
>        <primitive class= ocf provider= heartbeat type= drbd 
>id= drbd0 >
>
>          <instance_attributes id= ia-drbd0 >
>
>            <nvpair id= ia-drbd0-1 name= drbd_resource 
>value= drbd0 />
>
>            <nvpair id= ia-drbd0-2 name= clone_overrides_hostname 
>value= yes />
>
>          </instance_attributes>
>

>    <constraints>
>
>      <rsc_location id= location-ip-c001drbd01a rsc= ip-c001drbd01a >
>
>        <rule id= ip-c001drbd01a-rule score= -INFINITY >
>
>          <expression id= exp-ip-c001drbd01a-rule value= b 
>attribute= site operation= eq />
>
>        </rule>
>
>      </rsc_location>
>
>      <rsc_location id= location-ip-c001drbd01b rsc= ip-c001drbd01b >
>
>        <rule id= ip-c001drbd01b-rule score= -INFINITY >
>
>          <expression id= exp-ip-c001drbd01b-rule value= a 
>attribute= site operation= eq />
>
>        </rule>
>
>      </rsc_location>
>
>      <rsc_location id= drbd0-master-1 rsc= ms-drbd0 >
>
>        <rule id= drbd0-master-on-c001mlb_node01a role= master 
>score= 100 >
>
>          <expression id= expression-1 attribute= #uname 
>operation= eq value= c001mlb_node01a />
>
>        </rule>
>
>      </rsc_location>
>
>      <rsc_order id= order-drbd0-after-ip-c001drbd01a 
>first= ip-c001drbd01a then= ms-drbd0 score= 1 />
>
>      <rsc_order id= order-drbd0-after-ip-c001drbd01b 
>first= ip-c001drbd01b then= ms-drbd0 score= 1 />
>
>      <rsc_colocation rsc= ip-c001drbd01a score= INFINITY 
>id= colocate-drbd0-ip-c001drbd01a with-rsc= ms-drbd0 />
>
>      <rsc_colocation rsc= ip-c001drbd01b score= INFINITY 
>id= colocate-drbd0-ip-c001drbd01b with-rsc= ms-drbd0 />
>
>    </constraints>
>
>  </configuration>
*snip*

>ip-c001drbd01a (ocf::heartbeat:IPaddr2):      Started c001mlb_node01a
>
>Master/Slave Set: ms-drbd0
>
>        Stopped: [ drbd0:0 drbd0:1 drbd0:2 drbd0:3 drbd0:4 drbd0:5
>drbd0:6 drbd0:7 drbd0:8 drbd0:9 drbd0:10 drbd0:11 drbd0:12 drbd0:13
>drbd0:14 drbd0:15 drbd0:1
>6 drbd0:17 drbd0:18 drbd0:19 drbd0:20 drbd0:21 drbd0:22 drbd0:23
>drbd0:24 drbd0:25 ]
too many clones, see above.

->
>May 26 19:14:48 c001mlb_node01a cib: [31521]: info: retrieveCib: 
>Reading cluster configuration from: /var/lib/heartbeat/crm/cib.2Ij02L (digest:
>/var/lib/heartbeat/crm/cib.G64hBN)
>
>May 26 19:14:48 c001mlb_node01a pengine: [25734]: notice: print_list: 
>Stopped: [ drbd0:0 drbd0:1 drbd0:2 drbd0:3 drbd0:4 drbd0:5 drbd0:6
>drbd0:7 drbd0:8 drbd0:9 drbd0:10 drbd0:11 drbd0:12 drbd0:13 drbd0:14
>drbd0:15 drbd0:16 drbd0:17 drbd0:18 drbd0:19 drbd0:20 drbd0:21 
>drbd0:22
>drbd0:23 drbd0:24 drbd0:25 ]
all resources stopped. seems ok.

*snip*

>May 26 19:14:48 c001mlb_node01a pengine: [25734]: info: master_color:
>Promoting drbd0:0 (Stopped c001mlb_node01a)
>
>May 26 19:14:48 c001mlb_node01a pengine: [25734]: info: master_color:
>ms-drbd0: Promoted 1 instances of a possible 1 to master
>
>May 26 19:14:48 c001mlb_node01a pengine: [25734]: info: master_color:
>ms-drbd0: Promoted 1 instances of a possible 1 to master

2 instances have been promoted. that is correct.

>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info:
>do_state_transition: State transition S_POLICY_ENGINE ->
>S_TRANSITION_ENGINE [ input=I_PE_SUCCESS cause=C_IPC_MESSAGE 
>origin=handle_response ]
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: ERROR: do_fsa_action:
>Action A_DC_TIMER_STOP took 812418818s to complete
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: ERROR: do_fsa_action:
>Action A_INTEGRATE_TIMER_STOP took 812418818s to complete
>
>May 26 19:14:48 c001mlb_node01a pengine: [25734]: WARN:
>process_pe_message: Transition 7: WARNINGs found during PE processing.
>PEngine Input stored in: /var/lib/pengine/pe-warn-771.bz2
> 
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: ERROR: do_fsa_action:
>Action A_FINALIZE_TIMER_STOP took 812418818s to complete
>
>May 26 19:14:48 c001mlb_node01a pengine: [25734]: info:
>process_pe_message: Configuration WARNINGs found during PE processing. 
>Please run  crm_verify -L to identify issues.

warnings found - please see this pe-warn file and/or run crm_verify -L

>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: unpack_graph:
>Unpacked transition 7: 17 actions in 17 synapses
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: do_te_invoke:
>Processing graph 7 (ref=pe_calc-dc-1243365288-55) derived from
>/var/lib/pengine/pe-warn-771.bz2
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: ERROR: do_fsa_action:
>Action A_TE_INVOKE took 812418817s to complete
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_rsc_command:
>Initiating action 5: start ip-c001drbd01a_start_0 on c001mlb_node01a 
>(local)
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: do_lrm_rsc_op:
>Performing key=5:7:0:c58a32ec-ae57-4bc8-8a1e-5d7069c2f2bd
>op=ip-c001drbd01a_start_0 )

drbd seems to be up, starting ip below

>May 26 19:14:48 c001mlb_node01a lrmd: [25732]: info: rsc:ip-c001drbd01a:
>start
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_pseudo_action:
>Pseudo action 14 fired and confirmed
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_pseudo_action:
>Pseudo action 15 fired and confirmed
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_pseudo_action:
>Pseudo action 16 fired and confirmed
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_pseudo_action:
>Pseudo action 25 fired and confirmed
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_pseudo_action:
>Pseudo action 28 fired and confirmed
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_rsc_command:
>Initiating action 141: notify drbd0:0_post_notify_start_0 on 
>c001mlb_node01a (local)
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: do_lrm_rsc_op:
>Performing key=141:7:0:c58a32ec-ae57-4bc8-8a1e-5d7069c2f2bd
>op=drbd0:0_notify_0 )
>
>May 26 19:14:48 c001mlb_node01a lrmd: [25732]: info: rsc:drbd0:0: 
>notify
>
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: te_rsc_command:
>Initiating action 143: notify drbd0:0_post_notify_promote_0 on 
>c001mlb_node01a (local)
>
>May 26 19:14:48 c001mlb_node01a lrmd: [25732]: info: RA output:
>(ip-c001drbd01a:start:stderr) eth0:0: warning: name may be invalid
> 
>May 26 19:14:48 c001mlb_node01a crmd: [25735]: info: do_lrm_rsc_op:
>Performing key=143:7:0:c58a32ec-ae57-4bc8-8a1e-5d7069c2f2bd
>op=drbd0:0_notify_0 )
>
>May 26 19:14:48 c001mlb_node01a lrmd: [25732]: info: RA output:
>(drbd0:0:notify:stderr) 2009/05/26_19:14:48 INFO: drbd0: Using 
>hostname node_0
>
>May 26 19:14:48 c001mlb_node01a lrmd: [25732]: info: RA output:
>(ip-c001drbd01a:start:stderr) 2009/05/26_19:14:48 INFO: ip -f inet 
>addr add 192.168.80.213/32 brd 192.168.80.213 dev eth0 label eth0:0
>
>May 26 19:14:48 c001mlb_node01a lrmd: [25732]: info: RA output:
>(ip-c001drbd01a:start:stderr) 2009/05/26_19:14:48 INFO: ip link set 
>eth0 up 2009/05/26_19:14:48 INFO: /usr/lib64/heartbeat/send_arp -i 200 
>-r 5 -p /var/run/heartbeat/rsctmp/send_arp/send_arp-192.168.80.213 
>eth0
>192.168.80.213 auto not_used not_used

error starting ip.

as far as i can see, this is now repeating for different nodes and as pacemaker tries to recover.

please fix the above suggestions and reply with new findings.

i usually log to syslog so that i also see the drbd/daemon/... messages to aid with debugging.

moreover, please attach the files from now on and do not c/p them into one big email. it is a lot of work to read your email...

cheers,
raoul
[1] http://www.gossamer-threads.com/lists/linuxha/dev/50929
[2] http://www.clusterlabs.org/wiki/DRBD_HowTo_1.0
--
____________________________________________________________________
DI (FH) Raoul Bhatia M.Sc.         email.         r.bhatia at ipax.at
Technischer Leiter

IPAX - Aloy Bhatia Hava OEG        web.         http://www.ipax.at
Barawitzkagasse 10/2/2/11          email.           office at ipax.at
1190 Wien                          tel.              +43 1 3670030
FN 277995t HG Wien                 fax.           +43 1 3670030 15
____________________________________________________________________

_______________________________________________
Pacemaker mailing list
Pacemaker at oss.clusterlabs.org
http://oss.clusterlabs.org/mailman/listinfo/pacemaker

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20090527/738bd009/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: current-cib.xml
Type: application/octet-stream
Size: 20410 bytes
Desc: not available
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20090527/738bd009/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: drbf.conf
Type: application/octet-stream
Size: 976 bytes
Desc: not available
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20090527/738bd009/attachment-0003.obj>