On Oct 22, 2008, at 8:49 PM, Serge Dubrouski wrote:
It doesn't work for me either. Per Andrew pingd is probably still
broken and doesn't publish attributes.
I just checked again, and its definitely broken in 0.6 and 1.0
pingd is fine, but attrd (in attrd_local_callback) thinks that it has
already set that value in the cib and doesn't bother updating it.
the logic there is clearly wrong - shouldn't take long to fix.
On Wed, Oct 22, 2008 at 12:40 PM, Doug Huff <[EMAIL PROTECTED]>
wrote:
This isn't working in 1.0.0, either. I'm not seeing anything get
updated in
the transient_attributes section of either of the nodes in my cluster
configured almost identically to Andrew. Only difference I can tell
is
names. Here's the output from pingd -UV:
$ sudo /usr/lib/heartbeat/pingd -VU -a pingd -d 5s -m 1000 -h
10.10.2.1
pingd[26316]: 2008/10/22_13:37:54 debug: main: Adding ping host
10.10.2.1
pingd[26316]: 2008/10/22_13:37:54 info: main: Starting pingd
pingd[26316]: 2008/10/22_13:37:54 debug: stand_alone_ping: Checking
connectivity
pingd[26316]: 2008/10/22_13:37:54 debug: ping_open: Got address
10.10.2.1
for 10.10.2.1
pingd[26316]: 2008/10/22_13:37:54 debug: ping_open: Opened
connection to
10.10.2.1
pingd[26316]: 2008/10/22_13:37:54 debug: ping_write: Sent 39 bytes to
10.10.2.1
pingd[26316]: 2008/10/22_13:37:54 debug: dump_v4_echo: Echo from
10.10.2.1
(exp=1, seq=1, id=26316, dest=10.10.2.1, data=pingd-v4): Echo Reply
pingd[26316]: 2008/10/22_13:37:54 debug: dump_v4_echo: 59 bytes from
10.10.2.1, icmp_seq=1: pingd-v4
pingd[26316]: 2008/10/22_13:37:54 info: stand_alone_ping: Node
10.10.2.1 is
alive
pingd[26316]: 2008/10/22_13:37:54 debug: ping_close: Closed
connection to
10.10.2.1
pingd[26316]: 2008/10/22_13:37:54 info: send_update: 1 active ping
nodes
pingd[26316]: 2008/10/22_13:37:54 debug: log_data_element: Dumping
__FUNCTION__
pingd[26316]: 2008/10/22_13:37:54 info: log_data_element:
send_update: pingd
<__FUNCTION__ t="attrd" src="pingd" task="update" attr_name="pingd"
attr_value="1000" attr_dampening="5s" />
Here's the transient_attributes from both nodes:
<transient_attributes id="cf09b6b4-967e-4c7b-a1df-282d680d04f9">
<instance_attributes id="status-cf09b6b4-967e-4c7b-
a1df-282d680d04f9">
<nvpair id="status-cf09b6b4-967e-4c7b-a1df-282d680d04f9-
probe_complete"
name="probe_complete" value="true"/>
</instance_attributes>
</transient_attributes>
<transient_attributes id="16ad7ac8-732b-4c63-87e6-ad1194488d92">
<instance_attributes id="status-16ad7ac8-732b-4c63-87e6-
ad1194488d92">
<nvpair id="status-16ad7ac8-732b-4c63-87e6-ad1194488d92-
probe_complete"
name="probe_complete" value="true"/>
<nvpair
id="status-16ad7ac8-732b-4c63-87e6-ad1194488d92-fail-count-ld-dev-
dpa"
name="fail-count-ld-dev-dpa" value="1"/>
<nvpair
id="status-16ad7ac8-732b-4c63-87e6-ad1194488d92-last-failure-ld-dev-
dpa"
name="last-failure-ld-dev-dpa" value="1224688428"/>
</instance_attributes>
</transient_attributes>
_______________________________________________
Pacemaker mailing list
[email protected]
http://list.clusterlabs.org/mailman/listinfo/pacemaker
--
Serge Dubrouski.
_______________________________________________
Pacemaker mailing list
[email protected]
http://list.clusterlabs.org/mailman/listinfo/pacemaker
_______________________________________________
Pacemaker mailing list
[email protected]
http://list.clusterlabs.org/mailman/listinfo/pacemaker
