Mailing List Archive: crm_mon and last-failure timestamp not always present

crm_mon and last-failure timestamp not always present

Jan 7, 2015, 1:21 AM

Post #1 of 3 (1092 views)

Hello,
I have two old SLES 11 SP2 clusters, each one composed by two nodes, that
I'm comparing.

They should have same package versions (I' going to verify better this
point and configuration files when access will be granted to me) but on one
where there is a group configured with mysqld, when there is a failure I
get something like this:

/usr/sbin/crm_mon -1 --failcounts
...
Migration summary:
* Node sles1:
mysqld: migration-threshold=3 fail-count=1 last-failure='Tue Sep 16
10:38:37 2014'
* Node sles2:

On the other one where Novell eDir88 IDM service is present, I get
something like this with the same command:

Migration summary:
* Node srv-02-coll:
* Node srv-01-coll:
eDirectory: migration-threshold=3 fail-count=2

Is there any parameter inside configuration of cluster and/or resources
that could control if "last-failure" information will be shown or not?

Thanks in advance,

Gianluca

Re: crm_mon and last-failure timestamp not always present [ In reply to ]

gianluca.cecchi at gmail

Jan 7, 2015, 3:14 AM

Post #2 of 3 (1075 views)

Permalink

On Wed, Jan 7, 2015 at 10:21 AM, Gianluca Cecchi <gianluca.cecchi@gmail.com>
wrote:
[snip]

>
> Is there any parameter inside configuration of cluster and/or resources
> that could control if "last-failure" information will be shown or not?
>
>
>
Ok, solved.
On the cluster where timestamp is shown the resource has the meta parameter

meta failure-timeout="2min"

On the other cluster there isn't this kind of parameter.
I also tried to dynamically put and omit the parameter on a resource with 1
failure and the timestamp is shown or not, depending on the setting.
I think it is all related to this page features:
http://clusterlabs.org/doc/en-US/Pacemaker/1.0/html/Pacemaker_Explained/s-failure-migration.html

Gianluca

Re: crm_mon and last-failure timestamp not always present [ In reply to ]

emi2fast at gmail

Jan 7, 2015, 4:58 AM

Post #3 of 3 (1072 views)

Permalink

if you want to see the time stamp of your resource operations without
failure-timeout, because is used for reset the failcount, you can use
the "-t" option:

crm_mon -Arf1t

..........
..........

Operations:
* Node node01:
sambaip: migration-threshold=1000000
+ (30) start: rc=0 (ok)
Dummy1: migration-threshold=1000000 fail-count=2
+ (67) start: last-rc-change='Wed Jan 7 13:53:41 2015'
last-run='Wed Jan 7 13:53:41 2015' exec-time=20ms queue-time=0ms rc=0
(ok)
+ (68) monitor: interval=30000ms last-rc-change='Wed Jan 7
13:53:41 2015' exec-time=20ms queue-time=0ms rc=0 (ok)
Dummy2: migration-threshold=1000000
+ (69) start: last-rc-change='Wed Jan 7 13:53:41 2015'
last-run='Wed Jan 7 13:53:41 2015' exec-time=20ms queue-time=0ms rc=0
(ok)
+ (70) monitor: interval=2400000ms last-rc-change='Wed Jan 7
13:53:41 2015' exec-time=20ms queue-time=0ms rc=0 (ok)
Dummy3: migration-threshold=1000000
+ (71) start: last-rc-change='Wed Jan 7 13:53:41 2015'
last-run='Wed Jan 7 13:53:41 2015' exec-time=20ms queue-time=0ms rc=0
(ok)
+ (72) monitor: interval=2400000ms last-rc-change='Wed Jan 7
13:53:41 2015' exec-time=10ms queue-time=0ms rc=0 (ok)

2015-01-07 12:14 GMT+01:00 Gianluca Cecchi <gianluca.cecchi@gmail.com>:
> On Wed, Jan 7, 2015 at 10:21 AM, Gianluca Cecchi <gianluca.cecchi@gmail.com>
> wrote:
> [snip]
>
>>
>>
>> Is there any parameter inside configuration of cluster and/or resources
>> that could control if "last-failure" information will be shown or not?
>>
>>
>
> Ok, solved.
> On the cluster where timestamp is shown the resource has the meta parameter
>
> meta failure-timeout="2min"
>
> On the other cluster there isn't this kind of parameter.
> I also tried to dynamically put and omit the parameter on a resource with 1
> failure and the timestamp is shown or not, depending on the setting.
> I think it is all related to this page features:
> http://clusterlabs.org/doc/en-US/Pacemaker/1.0/html/Pacemaker_Explained/s-failure-migration.html
>
> Gianluca
>
>
> _______________________________________________
> Pacemaker mailing list: Pacemaker@oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
>

--
esta es mi vida e me la vivo hasta que dios quiera

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

Project Home: http://www.clusterlabs.org
Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
Bugs: http://bugs.clusterlabs.org