Active IQ Unified Manager Discussions

Grafana Harvest OCUM Couldn't execute monitoring template

IMHOTEPSON
6,123 Views

we always see this message in Hostname_netapp-harvest.log

 

[2017-07-03 13:19:28] [NORMAL ] WORKER STARTED [Version: 1.3] [Conf: netapp-harvest.conf] [Poller: OCUMSERVER_FQDN]
[2017-07-03 13:19:28] [NORMAL ] [main] Poller will monitor a [OCUM] at [OCUMSERVER_FQDN:443]
[2017-07-03 13:19:28] [NORMAL ] [main] Poller will use [password] authentication with username [svc_harvest] and password [**********]
[2017-07-03 13:19:29] [WARNING] [sysinfo] Discovered [SOME_UNWANTED_CLUSTERNAME] on OCUM server with no matching conf section; to collect this cluster please add a section
[2017-07-03 13:19:29] [WARNING] [sysinfo] Discovered [SOME_UNWANTED_CLUSTERNAME] on OCUM server with no matching conf section; to collect this cluster please add a section
[2017-07-03 13:19:29] [NORMAL ] [main] Collection of system info from [OCUMSERVER_FQDN] running [7.1] successful.
[2017-07-03 13:19:29] [ERROR  ] [main] Couldn't execute monitoring template [/opt/netapp-harvest/template/ocum-opm-hierachy.conf]: No such file or directory

 

and we do not see any data of the configured filers, and we do have the ocum-opm-hierarchy.conf at the shown path (copied from examples)

 

drwxr-xr-x 2 root root 4096 Nov 10  2016 example
drwxr-xr-x 2 root root 4096 Nov 10  2016 default
drwxr-xr-x 4 root root 4096 Jul  3 11:15 .
drwxr-xr-x 8 root root 4096 Jul  3 13:18 ..
-rw-r--r-- 1 root root 2892 Jul  3 13:18 ocum-opm-hierarchy.conf

 

netapp-harvest.conf

 

[OCUMSERVER_FQDN]
hostname          = OCUMSERVER_FQDN
group             = NETAPP_OCUM
host_type         = OCUM
data_update_freq  = 900
normalized_xfer   = gb_per_sec
username         = HARVEST_USERNAME
password         = PWD
host_port        = 443
template        = ocum-opm-hierachy.conf
graphite_root               = netapp-capacity.Clusters.{display_name}
graphite_meta_metrics_root  = netapp-capacity-poller.OCUMSERVER_FQDN


[cluster1]
hostname         = cluster1_FQDN
group             = NETAPP_OCUM
host_enabled = 0

[cluster2]
hostname         = cluster2_FQDN
group             = NETAPP_OCUM
host_enabled = 0

what do have todo to get the Perfdata collected from OCUM OCPM ??

1 ACCEPTED SOLUTION

madden
6,099 Views

Hi @IMHOTEPSON

 

 

Your poller entry has some issues in red:

 

[OCUMSERVER_FQDN]
hostname          = OCUMSERVER_FQDN
group             = NETAPP_OCUM
host_type         = OCUM
data_update_freq  = 900
normalized_xfer   = gb_per_sec
username         = HARVEST_USERNAME
password         = PWD
host_port        = 443
template        = ocum-opm-hierachy.conf
graphite_root               = netapp-capacity.Clusters.{display_name}
graphite_meta_metrics_root  = netapp-capacity-poller.OCUMSERVER_FQDN

1_ Update your FQDN to not have any '.' characters in it because these are used as a seperator by Graphite.  Maybe replace '.' with '_' or just use shortname?

2_ Template name has a misspelling; you are missing an 'r'

 

 

The 2_ is fatal, the 1_ is just a tip to improve.

 

Also, using a non-default template file like this is only recommended if you are using OPM external data provider for your perf metrics.  If you don't care about perf metrics, or will use Harvest for them, I would just use the default OCUM hierarchy.  In this way at least the Harvest provided dashboard capacity info will be of some use to you, otherwise you will have to create these from sratch.  To try this remove the last three lines in the config above.  

 

 

Hope this helps!

 

Cheers,
Chris Madden

Solution Architect - 3rd Platform - Systems Engineering NetApp EMEA (and author of Harvest)

Blog: It all begins with data

 

If this post resolved your issue, please help others by selecting ACCEPT AS SOLUTION or adding a KUDO or both!

 

 

View solution in original post

5 REPLIES 5

madden
6,100 Views

Hi @IMHOTEPSON

 

 

Your poller entry has some issues in red:

 

[OCUMSERVER_FQDN]
hostname          = OCUMSERVER_FQDN
group             = NETAPP_OCUM
host_type         = OCUM
data_update_freq  = 900
normalized_xfer   = gb_per_sec
username         = HARVEST_USERNAME
password         = PWD
host_port        = 443
template        = ocum-opm-hierachy.conf
graphite_root               = netapp-capacity.Clusters.{display_name}
graphite_meta_metrics_root  = netapp-capacity-poller.OCUMSERVER_FQDN

1_ Update your FQDN to not have any '.' characters in it because these are used as a seperator by Graphite.  Maybe replace '.' with '_' or just use shortname?

2_ Template name has a misspelling; you are missing an 'r'

 

 

The 2_ is fatal, the 1_ is just a tip to improve.

 

Also, using a non-default template file like this is only recommended if you are using OPM external data provider for your perf metrics.  If you don't care about perf metrics, or will use Harvest for them, I would just use the default OCUM hierarchy.  In this way at least the Harvest provided dashboard capacity info will be of some use to you, otherwise you will have to create these from sratch.  To try this remove the last three lines in the config above.  

 

 

Hope this helps!

 

Cheers,
Chris Madden

Solution Architect - 3rd Platform - Systems Engineering NetApp EMEA (and author of Harvest)

Blog: It all begins with data

 

If this post resolved your issue, please help others by selecting ACCEPT AS SOLUTION or adding a KUDO or both!

 

 

IMHOTEPSON
6,095 Views

fixed, we have checked the configuration and replaced the templatename, error is not displayed any more, but still get no data ;(

madden
6,089 Views

Polling frequency is every 15 minutes (00, 15, 30, 45 past the hour).

 

In you original log you had 

 

 

 

[2017-07-03 13:19:29] [WARNING] [sysinfo] Discovered [SOME_UNWANTED_CLUSTERNAME] on OCUM server with no matching conf section; to collect this cluster please add a section
[2017-07-03 13:19:29] [WARNING] [sysinfo] Discovered [SOME_UNWANTED_CLUSTERNAME] on OCUM server with no matching conf section; to collect this cluster please add a section

 

This indicates two clusters were discovered in OCUM, neither of which are in your config file.  You need to add them if you want data for them.  If you want data for some other cluster you first need to add it to OCUM, to allow collection by Harvest

 

Hope this helps!

 

Cheers,
Chris Madden

Solution Architect - 3rd Platform - Systems Engineering NetApp EMEA (and author of Harvest)

Blog: It all begins with data

 

If this post resolved your issue, please help others by selecting ACCEPT AS SOLUTION or adding a KUDO or both!

IMHOTEPSON
6,062 Views

since we have fixed the config, we

- have deleted all the (whisper) data which we got from the direct filer connects

- restarted the harvest

 

declare use of log data: the two unwanted are 2 clusters we do not want, but this shows, we generally have connect to OCUM (that was the idea to post this info)

 

now we see data coming in on graphite, but still did not see anything, no site, no structure no data on dasboards ;(

 

looks like we need some Dashboards for OCUM delivered Data...?!

madden
6,056 Views

The harvest supplied dashboards assume you use the default collection templates for perf + capacity collection.  If you don't, then you have to make your own dashboards.  This is why I suggested to delete the last three lines from your OCUM poller entry in netapp-harvest.conf....if you do then you could use the capacity sections of the default dashboards provided by harvest.

 

Cheers,
Chris Madden

Solution Architect - 3rd Platform - Systems Engineering NetApp EMEA (and author of Harvest)

Blog: It all begins with data

 

If this post resolved your issue, please help others by selecting ACCEPT AS SOLUTION or adding a KUDO or both!

Public