Ask Your Question

No memory and disk.ephemeral.size usage meters from ceilometer

asked 2016-02-17 22:18:26 -0500

Lee gravatar image

updated 2016-02-18 01:16:34 -0500

Hi everybody.

I have installed openstack juno ubuntu 14.04 on all nodes and trying to get the memory and disk.ephemeral.size. However, on checking the "ceilometer meter-list" I can't find memory or memory.usage or disk.ephemeral.size meters.

When i checked the meter using "ceilomer meter-list", currently the output of ceilometer meter-list is as follows: cpu, cpu_util,,, disk.device.write.bytes, disk.device.write.requests,,,,, disk.write.bytes, disk.write.bytes.rate, disk.write.requests, disk.write.requests.rate, image, image.size, instance, network.incoming.bytes, network.incoming.bytes.rate, network.incoming.packets, network.incoming.packets.rate, network.outgoing.bytes, network.outgoing.bytes.rate, network.outgoing.packets, network.outgoing.packets.rate

I want to get the data (memory, memory.usage, and disk.ephemeral.size), why i am not able to get other meters data?

Please provide a solution for this.

Thanks and Regards!

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2016-02-18 01:57:34 -0500

yprokule gravatar image

Hi Lee,

Both 'memory' and ' disk.ephemeral.size' are derived from notifications emitted by Nova. Please check if your compute is configured to send the notifications.

Regarding 'memory.usage' I advise to check ceilometer/compute.log

edit flag offensive delete link more


Currently, I try to check the mongodb.log and ceilometer.log in controller node. The compute is sending the 'memory' and 'disk.ephemeral.size' query of resource. However, mongodb is not updated in real time. I am going to check frequently the mongodb for ceilometer. Thanks yprokule !

Lee gravatar imageLee ( 2016-02-19 02:10:57 -0500 )edit

@yprokule Thanks! Thanks for the info. However, I do not search the both 'memory' and ' disk.ephemeral.size'. So, do i reconfigure nova.conf of compute or ceilometer.conf of compute?


Lee gravatar imageLee ( 2016-02-21 23:16:55 -0500 )edit

@Lee, IIRC U have to configure nova.conf on the controller node. Regarding 'memory.usage' : - meter 'memory.usage' has to be configured in the pipeline.yaml on the compute node and controller - Check notes

yprokule gravatar imageyprokule ( 2016-02-23 04:53:18 -0500 )edit

@yprokule Thanks for the response yprokule. I have not solved this issue. can you share me the pipeline.yaml file? I want to refer your pipeline.yaml file. Thanks!

Lee gravatar imageLee ( 2016-02-28 23:27:16 -0500 )edit

Add this snippet to your pipeline.yaml at compute and controller nodes and restart ceilometer services: - name: mem_source interval: 600 meters: - "memory.usage" - "memory.resident" sinks: - meter_sink

Then check ceilometer/compute.log for additional info

yprokule gravatar imageyprokule ( 2016-02-28 23:52:50 -0500 )edit

@yprokule Thanks for the response yprokule. This is my pipeline.yaml code. - Check notes (

You mean the pipeline.yaml has to revise it? - name: mem_source interval: 600 meters: - "memory.usage" - "memory.resident"

Lee gravatar imageLee ( 2016-02-29 00:46:44 -0500 )edit

@yprokule I modified my pipeline.yaml at compute and controller nodes. However, I was confirmed that an error had occurred.

Lee gravatar imageLee ( 2016-03-01 18:21:56 -0500 )edit

@Lee, pipeline.yaml U've sent is ok. What error do U see? Do U see that memory* pollsters are run on the compute node?

yprokule gravatar imageyprokule ( 2016-03-02 00:43:05 -0500 )edit

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2016-02-17 22:18:26 -0500

Seen: 954 times

Last updated: Feb 18 '16