[Pacemaker] Pacemaker 0.7.3: How to use pingd clone resource and constraints

Bruno Voigt Bruno.Voigt at ic3s.de
Mon Oct 6 22:41:44 UTC 2008


Andrew Beekhof wrote:
>> In short - pingd in 0.7 is/was completely broken on linux (but worked
>> great on Darwin).
>> Sorry :(
>>
>> The good news is that I have it basically functioning properly now - I
>> just need to clean up the patch before I commit (which I'll do over
>> the weekend).
>>
>>     
>
> Would you be able to verify that pingd in the new packages (0.7.3-2.1)
> behaves properly?
>   
I installed the debian/etch packages from Oct 06 and no longer see "bad
echo" messages:

Oct  7 00:32:33 xen20b attrd_updater: [29206]: info: Invoked:
/usr/sbin/attrd_updater -S status -n cpu_speed -v 4002.38
Oct  7 00:32:33 xen20b attrd_updater: [29209]: info: Invoked:
/usr/sbin/attrd_updater -S status -n cpu_cores -v 2
Oct  7 00:32:33 xen20b attrd_updater: [29214]: info: Invoked:
/usr/sbin/attrd_updater -S status -n cpu_load -v 2.05
Oct  7 00:32:33 xen20b attrd_updater: [29226]: info: Invoked:
/usr/sbin/attrd_updater -S status -n ram_total -v 550
Oct  7 00:32:33 xen20b attrd_updater: [29236]: info: Invoked:
/usr/sbin/attrd_updater -S status -n ram_free -v 100
Oct  7 00:32:33 xen20b attrd_updater: [29246]: info: Invoked:
/usr/sbin/attrd_updater -S status -n root_free -v 40%
Oct  7 00:32:36 xen20b pingd: [12186]: info: stand_alone_ping: Node
195.244.97.241 is alive
Oct  7 00:32:36 xen20b pingd: [12186]: info: send_update: 1 active ping
nodes
Oct  7 00:32:42 xen20b pingd: [12197]: info: stand_alone_ping: Node
172.17.32.23 is alive
Oct  7 00:32:42 xen20b pingd: [12197]: info: stand_alone_ping: Node
192.168.132.23 is alive
Oct  7 00:32:42 xen20b pingd: [12197]: info: send_update: 2 active ping
nodes
Oct  7 00:32:46 xen20b pingd: [12186]: info: stand_alone_ping: Node
195.244.97.241 is alive
Oct  7 00:32:46 xen20b pingd: [12186]: info: send_update: 1 active ping
nodes
Oct  7 00:32:52 xen20b pingd: [12197]: info: stand_alone_ping: Node
172.17.32.23 is alive
Oct  7 00:32:52 xen20b pingd: [12197]: info: stand_alone_ping: Node
192.168.132.23 is alive
Oct  7 00:32:52 xen20b pingd: [12197]: info: send_update: 2 active ping
nodes
Oct  7 00:32:53 xen20b attrd: [3562]: info: attrd_ha_callback: (null)
message from (null)
Oct  7 00:32:53 xen20b attrd: [3562]: info: find_hash_entry: Ignoring
message with no attribute name
Oct  7 00:32:53 xen20b attrd_updater: [29337]: info: Invoked:
/usr/sbin/attrd_updater -S status -n arch -v x86_64
Oct  7 00:32:53 xen20b attrd_updater: [29340]: info: Invoked:
/usr/sbin/attrd_updater -S status -n os -v Linux-2.6.18-6-xen-amd64
Oct  7 00:32:53 xen20b attrd_updater: [29363]: info: Invoked:
/usr/sbin/attrd_updater -S status -n free_swap -v 2050
Oct  7 00:32:53 xen20b attrd_updater: [29370]: info: Invoked:
/usr/sbin/attrd_updater -S status -n cpu_info -v Intel(R) Xeon(R)
CPU            5110  @ 1.60GHz
Oct  7 00:32:53 xen20b attrd_updater: [29371]: info: Invoked:
/usr/sbin/attrd_updater -S status -n cpu_speed -v 4002.38
Oct  7 00:32:53 xen20b attrd_updater: [29372]: info: Invoked:
/usr/sbin/attrd_updater -S status -n cpu_cores -v 2
Oct  7 00:32:53 xen20b attrd_updater: [29377]: info: Invoked:
/usr/sbin/attrd_updater -S status -n cpu_load -v 2.05
Oct  7 00:32:53 xen20b attrd_updater: [29387]: info: Invoked:
/usr/sbin/attrd_updater -S status -n ram_total -v 550
Oct  7 00:32:53 xen20b attrd_updater: [29397]: info: Invoked:
/usr/sbin/attrd_updater -S status -n ram_free -v 100
Oct  7 00:32:53 xen20b attrd_updater: [29407]: info: Invoked:
/usr/sbin/attrd_updater -S status -n root_free -v 40%
Oct  7 00:32:56 xen20b pingd: [12186]: info: stand_alone_ping: Node
195.244.97.241 is alive
Oct  7 00:32:56 xen20b pingd: [12186]: info: send_update: 1 active ping
nodes

But I can't detect in the cibadmin -Q output any trace of the set
attributes!
Where should they be located ?
Do I have to initialize them manually so that pingd can update them? How?
BTW: the output of the SysInfo clone-resource is there:
 
 <node_state uname="xen20b.fqdn" ha="active" in_ccm="true" crmd="online"
join="member" shutdown="0" expected="member"
id="278bebc6-2a59-4fa9-be2f-f6e262ce8936"
crm-debug-origin="do_update_resource">
      <transient_attributes id="278bebc6-2a59-4fa9-be2f-f6e262ce8936">
        <instance_attributes
id="status-278bebc6-2a59-4fa9-be2f-f6e262ce8936">
          <nvpair name="root_free"
id="status-278bebc6-2a59-4fa9-be2f-f6e262ce8936-root_free" value="40%"/>
          <nvpair id="status-278bebc6-2a59-4fa9-be2f-f6e262ce8936-os"
name="os" value="Linux-2.6.18-6-xen-amd64"/>
          <nvpair id="status-278bebc6-2a59-4fa9-be2f-f6e262ce8936-arch"
name="arch" value="x86_64"/>

# ps -ef | grep pingd
root      6440     1  0 Oct06 ?        00:00:00 /usr/lib/heartbeat/pingd
-D -p /var/run/heartbeat/rsctmp/pingd-pingd-internal:0 -a pingd-internal
-d 5s -m 1000 -h 172.17.32.23 -h 192.168.132.23
root      6502     1  0 Oct06 ?        00:00:00 /usr/lib/heartbeat/pingd
-D -p /var/run/heartbeat/rsctmp/pingd-pingd-external:0 -a pingd-external
-d 5s -m 1000 -h 195.244.97.241
 
> Btw. You can also run it from the command line (with no cluster
> present) by supplying the -U option.
> Instead of trying to update attrd, it will instead output the message
> it would have sent.
>
>   
/usr/lib/heartbeat/pingd -U -a pingd-internal -d 5s -m 1000 -h 172.17.32.23
does not generate any output  on the console.
The -U switch does not appear in pingd --help.

WR,
Bruno

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20081007/a13a435b/attachment-0002.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: OpenPGP digital signature
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20081007/a13a435b/attachment-0002.sig>


More information about the Pacemaker mailing list