Hi,
On Wed, Dec 19, 2007 at 01:57:17PM +0900, Keisuke MORI wrote:
Dejan,
Thank you for commiting the SNMP extention to the repository, but
Would you please include the following patches along with it?
1) SNMP: fix a problem on displaying an unmanaged group and to check it
in
On Tue, Dec 18, 2007 at 11:14:05AM -0700, Serge Dubrouski wrote:
Dejan -
Did you cover this in the Mercurial?
Just now. Thanks for the reminder.
Dejan
On Dec 16, 2007 6:20 PM, HIDEO YAMAUCHI [EMAIL PROTECTED] wrote:
Hi,
Please disregard this. I found where was the problem and here
Hi,
I'm looking for a checklist, so we can simulate different errors.
Example:
- failure with switch at node 1
take out the network wire out of the switch and look if node 2 takes
the prozess
Does such a list exist?
There are so many failures that can happen, so that it would be nice to
On Dec 18, 2007, at 5:35 PM, Scott Mann wrote:
On Tue 12/18/2007 12:52 AM, Andrew Beekhof said:
On Dec 17, 2007, at 11:28 PM, Scott Mann wrote:
in v2 mode you can't monitor the resource using the HA API... only
via
the CIB.
Yes, right. Figured that out when my ha api call for
Hi,
On Wed, Dec 19, 2007 at 04:47:31PM +0900, HIDEO YAMAUCHI wrote:
Hi,
I installed a development version in the following procedures.
1)Initial installation of RHEL5(Update1).
2)Installation of libnet.
3)Installation of Heartbeat development version(Heartbeat-Dev-d739f7e38999).
Hi,
On Wed, Dec 19, 2007 at 08:52:00AM +0100, Jochen Lienhard wrote:
Hi,
I'm looking for a checklist, so we can simulate different errors.
Example:
- failure with switch at node 1
take out the network wire out of the switch and look if node 2 takes
the prozess
Does such a list
On Dec 19, 2007, at 8:47 AM, HIDEO YAMAUCHI wrote:
Hi,
I installed a development version in the following procedures.
1)Initial installation of RHEL5(Update1).
2)Installation of libnet.
3)Installation of Heartbeat development version(Heartbeat-Dev-
d739f7e38999).
4)Setting such as ha.cf
On Dec 18, 2007, at 5:49 AM, DAIKI MATSUDA wrote:
Hi, Andrew.
I am sorry for sending uncomplete information and attached the results
of 'cibadmin -Ql' for commonn cib.xml files.
So, could you confirm them?
i'm confused why do you think the group should move?
the rule rule id=loc1:rule1
On Dec 18, 2007, at 6:28 AM, DAIKI MATSUDA wrote:
Hi, Andrew.
I am sorry for delay because I was not aware of requiring logs. And I
attached the hb_report log.
Best Regards
MATSUDA, Daiki
2007/12/12, Andrew Beekhof [EMAIL PROTECTED]:
On Dec 12, 2007, at 7:41 AM, DAIKI MATSUDA wrote:
Hi,
Hi,
I am setting up a dedicated backend fileserver HA-cluster where I have
one dedicated network interface per server that needs to access the
data. This means that I am not really interrested in only doing failover
of secondary IP-addresses (or aliases), but rather the primary
IP-address of real
Hi,
I've a bit strange problem with cibadmin, I should replace actual
cib.xmlwith a new hand-edited
cib.xml, so I executed:
cibadmin -R -x /path/to/newfile/cib.xml, but next error is shown:
Call cib_replace failed (-45): Update was older than existing configuration
null
But Im sure it's not
Hi,
On Wed, Dec 19, 2007 at 11:34:38AM +0100, Fernando Iglesias wrote:
Hi,
I've a bit strange problem with cibadmin, I should replace actual
cib.xmlwith a new hand-edited
cib.xml, so I executed:
cibadmin -R -x /path/to/newfile/cib.xml, but next error is shown:
Call cib_replace failed
On 2007-12-19T11:32:12, Andrew Beekhof [EMAIL PROTECTED] wrote:
i prefer to use the crm respawn directive which disables the fast-fail
logic^.
when a non-transient problem like this occurs and heartbeat is started at
boot time (which is the normal thing to do), you have about 2s to identify
Dejan Muhamedagic wrote:
To sum up, I have eth2 dedicated to a specific IP on both cluster nodes
and I want heartbeat to do failover for it. Is there a way to do this
without ending up with eth2:0, with the current resource scripts?
Why do you need to move the real interface addresses? What's
Hi,
On Wed, Dec 19, 2007 at 01:09:00PM +0100, Anders Bruun Olsen wrote:
Hi,
I am setting up a dedicated backend fileserver HA-cluster where I have
one dedicated network interface per server that needs to access the
data. This means that I am not really interrested in only doing failover
of
[EMAIL PROTECTED] wrote:
By having 2 Ethernet ports on each machine, one for the heartbeat and
one as an active interface. If you are using a crossover for the
heartbeat, the active interface would allow you to access the machine
when it was not your active node.
I have eth0 connected to our
By having 2 Ethernet ports on each machine, one for the heartbeat and
one as an active interface. If you are using a crossover for the
heartbeat, the active interface would allow you to access the machine
when it was not your active node.
Simon
-Original Message-
From: [EMAIL
What we've done is set up our machines with crossover Ethernet on Eth1
for synch and has IP addresses for our machines on Eth0 and a floating
IP address for failover for our applications. This has been working in
our production environment for our front end (API) boxes on a MySQL
cluster
On Dec 19, 2007, at 2:13 PM, Lars Marowsky-Bree wrote:
On 2007-12-19T11:32:12, Andrew Beekhof [EMAIL PROTECTED] wrote:
i prefer to use the crm respawn directive which disables the fast-
fail
logic^.
when a non-transient problem like this occurs and heartbeat is
started at
boot time (which
Hi, I'm trying to put 2 sets of rules for stickiness and I'm having crm_mon
and crmd segfaulting if I try to put them.
I've narrowed it down to putting this rule :
cibadmin -C -o crm_config -X '
cluster_property_set id=heures_maintenance score=100
rule id=cluster:heures_nuit boolean_op=or
On Dec 19, 2007, at 3:10 PM, Cousin Marc wrote:
Hi, I'm trying to put 2 sets of rules for stickiness and I'm having
crm_mon
and crmd segfaulting if I try to put them.
I've narrowed it down to putting this rule :
cibadmin -C -o crm_config -X '
cluster_property_set id=heures_maintenance
sorry, I knew I was forgetting something :)
Version :
2.1.2+hg.11310.702e4f418ca8-2 (debian sid)
The bt from core dump for crm_mon (it's not compiled with debug, I guess...):
(gdb) bt full
#0 0xb7e4d35c in cron_range_satisfied () from /usr/lib/libpe_status.so.1
No symbol table info available.
On Dec 19, 2007, at 4:09 PM, Cousin Marc wrote:
sorry, I knew I was forgetting something :)
Version :
2.1.2+hg.11310.702e4f418ca8-2 (debian sid)
The bt from core dump for crm_mon (it's not compiled with debug, I
guess...):
nod :-(
but at least i have the exact version number - it's
On Dec 19, 2007, at 4:09 PM, Cousin Marc wrote:
sorry, I knew I was forgetting something :)
Version :
2.1.2+hg.11310.702e4f418ca8-2 (debian sid)
Ah, this was fixed back in October
http://hg.beekhof.net/lha/crm-dev/rev/64fd1fb9d725
If it helps, the patch is included in the latest interim
2007/12/19, Andrew Beekhof [EMAIL PROTECTED]:
On Dec 18, 2007, at 5:49 AM, DAIKI MATSUDA wrote:
Hi, Andrew.
I am sorry for sending uncomplete information and attached the results
of 'cibadmin -Ql' for commonn cib.xml files.
So, could you confirm them?
i'm confused why do you
2007/12/19, Andrew Beekhof [EMAIL PROTECTED]:
On Dec 18, 2007, at 6:28 AM, DAIKI MATSUDA wrote:
Hi, Andrew.
I am sorry for delay because I was not aware of requiring logs. And I
attached the hb_report log.
Best Regards
MATSUDA, Daiki
2007/12/12, Andrew Beekhof [EMAIL
Junko IKEDA wrote:
Thanks a lot!
Would you take in this fix to 2.1.3?
I hope it's not too late.
OK... It's a regression - and I don't think our tests catch this case,
so I doubt it will affect the outcome of tests. Nevertheless...
Dave, Dejan and (if possible) Lars:
I have put this patch
Hi,
While I check crm_mon display, it keeps node's status as standby after
that node was shutdown.
CIB status is here;
node_state
uname=prec370e crmd=offline
in_ccm=false ha=dead join=down
id=9d9ca527-cea9-470c-9e03-e49fe5630bba shutdown=0 expected=down
28 matches
Mail list logo
