Active IQ Unified Manager Discussions

NABox / Netapp-Harvest LUN Dashboard issue

JayBarker
4,206 Views

Hello,

 

I am having a strange issue with a fresh install of NABox 2.2 , we installed the appliance and configured it to talk to one of our C-Mode clusters and all was working well for half the day, all the dashboards working as expected. After about 5 hours the LUN and Volume dashboards no longer show any metrics, the highlights just show "N/A ms" across all metrics and all graphis show "No Datapoints" . I am very new to this so would appriciate some guidance on what logs to check to troubleshoot.

4 REPLIES 4

yannb
4,178 Views

Hi Jay,

 

You can you send me a support bundle at yann.bizeul@netapp.com I will take a look at it.

 

Usually where I start with that kind of issue is going in the virtual appliance shell and check Harvest logs in /opt/netapp-harvest/logs.

 

They are also exported in the support bundle if you create and open it.

 

A time change might be the source of that kind of problem as well.

JayBarker
4,138 Views

Thanks - send the bundle across.

 

I checked the logs at the location and nothing seems out of the ordinary -

 

root@nabox:~# tail -f /opt/netapp-harvest/log/NETAPPLRO_netapp-harvest.log
[2017-01-31 18:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1849, plugin_time=29, metrics=1064441, skips=0, fails=0
[2017-01-31 22:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1924, plugin_time=29, metrics=1098018, skips=0, fails=0
[2017-02-01 02:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1995, plugin_time=28, metrics=1011047, skips=0, fails=0
[2017-02-01 06:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1865, plugin_time=27, metrics=1005026, skips=0, fails=0
[2017-02-01 10:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1873, plugin_time=29, metrics=1046223, skips=0, fails=0
[2017-02-01 14:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1883, plugin_time=29, metrics=1084037, skips=0, fails=0
[2017-02-01 18:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1875, plugin_time=29, metrics=1056884, skips=0, fails=0
[2017-02-01 22:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1945, plugin_time=29, metrics=1092523, skips=0, fails=0
[2017-02-02 02:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=2009, plugin_time=28, metrics=1006703, skips=0, fails=0
[2017-02-02 06:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1881, plugin_time=28, metrics=996057, skips=0, fails=0
[2017-02-02 10:29:47] [NORMAL ] Poller status: status, secs=14400, api_time=1891                                                                                                                                                             , plugin_time=29, metrics=1048410, skips=0, fails=0

 

 

yannb
4,082 Views

Following up on that issue for our fellow members.

 

Some dashboard are not meant to be used with a lot of performance objects in the graphite database.

 

For example, when the Volume: detail dashboard is brought up, it will query Graphite with all your volume names in the HTTP request, this will break at some point.

 

The easiest thing to do to avoid that issue is to restrict the number of objects by selecting only a few volumes you're interested in.

 

Another option is to modify the dashboard template and place a custom value for "All", there is a settings for that in the settings for the dashboard template value. The custom value should be "*" to have the same behavior for the dashboard.

 

You might then experience slowness if really there is a lot of volume but it shouldn't fail.

rezkmekhael
3,885 Views

check the filesystem /opt size is it full or not ?

 

 

root@nabox:~# df -hP
Filesystem Size Used Avail Use% Mounted on
/dev/sda6 4.8G 3.0G 1.7G 65% /
udev 10M 0 10M 0% /dev
tmpfs 6.3G 8.5M 6.3G 1% /run
tmpfs 16G 0 16G 0% /dev/shm
tmpfs 5.0M 0 5.0M 0% /run/lock
tmpfs 16G 0 16G 0% /sys/fs/cgroup
/dev/sda1 88M 29M 52M 36% /boot
/dev/mapper/vg_data-lv_data 493G 122G 349G 26% /opt              ---> it is 100% full or not

 

Public