thanks, sounds great

will try tomorrow. 

On 2017-12-21 at 6:26 PM, Pducharme said:

 

You can probably use the cadvisor (from Google) Docker to monitor ressources usage.

 

This is exactly what I was wondering- why this over cAdvisor or the other way around?

I noticed today that my netdata docker log is full of warnings-

--- BEGIN TRACE ---
--- END TRACE ---
2018-01-23 16:54:33: charts.d: : apcupsd: command 'apcaccess status 127.0.0.1:3551 ' failed:

--- BEGIN TRACE ---
--- END TRACE ---
2018-01-23 16:55:23: charts.d: : apcupsd: command 'apcaccess status 127.0.0.1:3551 ' failed:

--- BEGIN TRACE ---
--- END TRACE ---
2018-01-23 16:59:53: charts.d: : apcupsd: command 'apcaccess status 127.0.0.1:3551 ' failed:

--- BEGIN TRACE ---
--- END TRACE ---
2018-01-23 17:02:03: charts.d: : apcupsd: command 'apcaccess status 127.0.0.1:3551 ' failed:

--- BEGIN TRACE ---
--- END TRACE ---
2018-01-23 17:09:23: charts.d: : apcupsd: command 'apcaccess status 127.0.0.1:3551 ' failed:

--- BEGIN TRACE ---
--- END TRACE ---
2018-01-23 17:12:15: netdata ERROR : PLUGIN[proc]: File /var/cache/netdata/disk_qops.sda/main.db is too old. Clearing it.

Not quite sure what they mean? The apcupsd data is all being displayed in the docker web page.

Why is `/var/run/docker.sock` suggested to be mounted with write access?

On 8/14/2017 at 3:27 PM, BillClinton said:

 

 

hi Lappen71

Do you see the side menu on the right of the screen?

 

 

That is where the application monitoring is. It lists the current running dockers. Either with the docker ID string or the name if you have linked up the netdata docker right.

In each application list you'll see how much CPU each application is using.

 

 

Hope this helps

 

 

Sent from my iPad using Tapatalk

 

I don't have that list and I have apps running? how can I add them?

 

On the other hand how can I add other charts? like the sensors pluging?

https://github.com/firehol/netdata/wiki/Add-more-charts-to-netdata

 

I would like to see temperatures and voltajes

 

On 2/10/2018 at 9:59 AM, realies said:

Why is `/var/run/docker.sock` suggested to be mounted with write access?

Pretty sure netdata needs write access to request docker container names on the docker socket.

On 3/25/2018 at 6:49 PM, primeval_god said:

Pretty sure netdata needs write access to request docker container names on the docker socket.

 

Not true. @L0rdRaiden, map the container `/var/run/docker.sock` file to the host. Read-only would suffice.

1 hour ago, realies said:

 

Not true. @L0rdRaiden, map the container `/var/run/docker.sock` file to the host. Read-only would suffice.

Thanks, it works now, I think it just required some time to load, is a pitty that I see the codes but not the name of the application.

You should be able to see the name of the containers.

5 hours ago, realies said:

You should be able to see the name of the containers.

Like this? it doesn't work

Edited March 27, 20188 yr by L0rdRaiden

4 hours ago, L0rdRaiden said:

Like this? it doesn't work

 

 

Following the info in the first post set up a new path in the docker setup page like this:

 

Edited March 27, 20188 yr by BillClinton
Added info

17 minutes ago, BillClinton said:

 

 

Following the info in the first post set up a new path in the docker setup page like this:

 

 

I can confirm this works, though it might take a minute or two for the names to appear.

My log is spammed with this error:

 

Quote

2018-03-27 13:53:50: python.d ERROR: cpuidle: cpuidle: update() unhandled exception: [Errno 2] No such file or directory: '/host/proc/schedstat'

 

Any ideas?

 

Edited March 27, 20188 yr by SirBindy
added picture

20 hours ago, L0rdRaiden said:

Like this? it doesn't work

 

 

@roland, can we have the `/var/run/docker.sock` read-only mapping in the XML template please?

 

@SirBindy, your folder mapping is bad. Make it

 

2 hours ago, realies said:

@roland, can we have the `/var/run/docker.sock` read-only mapping in the XML template please?

Done !

 

Thanks @realies for pointing out the obvious.

GitHub is updated. https://github.com/Data-Monkey/docker-templates/blob/master/Data-Monkey/netdata.xml

 

 

On 24/03/2018 at 10:26 PM, L0rdRaiden said:

On the other hand how can I add other charts? like the sensors pluging?

https://github.com/firehol/netdata/wiki/Add-more-charts-to-netdata

Hi @L0rdRaiden,

 

Sorry, I have been missing for a while.

THere has been some development to the docker by titpetric over at dockerhub.

From his page:  https://hub.docker.com/r/titpetric/netdata/

Adding custom alarms, charts and configuration overrides

To add custom alarms, charts or to override any default configuration file, mount a volume to the container to /etc/netdata/override, like -v /opt/netdata/overrides:/etc/netdata/override. Then, place your config files in the directory as if it was /etc/netdata/.

For example to create a custom alarm for system temperature, create a health.d folder in your local directory (/opt/netdata/override in the example above) and place a sensors.conf file with your alarm configuration inside the health.d directory.

 

By default my template maps /etc/netdata to your config directory, in my case /mnt/cache/appdata/netdata

inside there you should find the override folder. Most likely empty. But that is where you can add your configurations.

I have not tried this yet, the things you mention show up in my config by default.

 

@realies my mappings match your suggestions exactly...

Edited March 29, 20188 yr by SirBindy

In the docker containers, this 2 graphs produces a perfect match (talking about the netdata screeshot), which makes sense, but in the VM I get what you can see in the picture bellow and it doesn't match with what unraid is reporting (unraid reported and average of 15-20% as much load during the period 10:50 and 10:55, and the CPU report included in the application (SophosXG) running in the VM was reporting almost a iddle load...

Why is this? I don't know what to trust.

 

I have all the cores assigned to SophosXG and looking at the netdata graphs everything else was mostly idle.

 

 

In te doc

Edited March 29, 20188 yr by L0rdRaiden

Is this still working?

 

I never met a app which throws out so many errors on startup, wow.

 

File not found, reports failure, module is needed, child exited with code 1... i guess thats every error on the world inside this container..

The "tutorial" didnt showed to change anythign... so why its not working?

 

Edit:
Okay, i got it working, just needed to restart (?) Maybe you could add that to the first page. I couldnt access it until i restartet it.

Edited April 3, 20188 yr by nuhll

On ‎2018‎-‎03‎-‎27 at 1:57 PM, SirBindy said:

 

 

 

I Have the same errors here. The /host/proc/schedstat file doesnt exist on my system. I was wondering if I need to edit the config file and remove the netdata.runtime_cpuidle to get ride of this spamming message. Any suggestion ? Thx

Hey... So I love netdata.  But I have a question about one of the stats it gives. 

As per the picture below you can see that "nice" is eating up a lot of my CPU usage.  What is nice? 

 

I accepted a merge request https://github.com/Data-Monkey/docker-templates/pull/3

to change the default config mapping to /etc/ netdata /override

 

From the merge request, thanks a lot acrelle:

Change the default netdata configuration to be /etc/netdata/override.

Scripts placed in here are automatically copied into /etc/netdata during the container start script.

/etc/netdata/override is the volume exposed in the original Dockerfile:

https://hub.docker.com/r/titpetric/netdata/~/dockerfile/

 

On 4/18/2018 at 7:56 PM, igowas said:

 

 

I Have the same errors here. The /host/proc/schedstat file doesnt exist on my system. I was wondering if I need to edit the config file and remove the netdata.runtime_cpuidle to get ride of this spamming message. Any suggestion ? Thx

I get that error too.

The application seems to run OK, but as long as it spits errors in log I think something might be wrong.

Can someone point me in the right direction to make function the NUT plugin? i've already the NUT plugin working in Unraid but i would greatly appreciate to monitor my ups also in netdata. but I can't figure how to do it :'(

 

On 6/18/2018 at 9:36 AM, playmake said:

Can someone point me in the right direction to make function the NUT plugin? i've already the NUT plugin working in Unraid but i would greatly appreciate to monitor my ups also in netdata. but I can't figure how to do it 

 

Netdata's existing NUT module will not work when Netdata is running inside a docker as in unRAID. I did come up with an alternative that has been working for me however. See the post i made on the previous page of this topic.