Monitoring filesystems in AWS EC2 instances (OA TEAM)

[Eduard Querol]

Hello,

We are trying to monitor metrics collected by cloudwatch agent with generic aspect. Attached you can find those metrics provided by aws cli list-metrics command.
We need monitoring filesystem metrics and to achive that we have added the dimension "path" to the file /var/opt/OV/conf/cloudmp/AWS/AWS_Generic_Discovery_Configuration.cfg:

"CWAgent":"InstanceId,path",

The problem is that metrics have not information about the EC2 instance that it belongs so in case of alert we cannot know the instance afeccted:

Example of filesystem metrics:
=
11/10/20 23:40:06
relatedCi
aws:arn:cwagent:eu-west-1:315117368111:path//tmp

11/10/20 23:40:06
originalName
path//tmp

11/10/20 23:40:06
metricClass
CWAgent_METRICS

11/10/20 23:40:06
unit
N/A

11/10/20 23:40:06
disk_inodes_free
496169.00

11/10/20 23:40:06
timeMeasured
1605048006

11/10/20 23:40:06
timeLogged
N/A

11/10/20 23:40:06
integrationId
N/A

11/10/20 23:40:06
node
eu-west-1

11/10/20 23:40:06
disk_used_percent
0.00

11/10/20 23:40:06
diskio_io_time
0.00

11/10/20 23:40:06
mem_used_percent
0.00

11/10/20 23:40:06
swap_used_percent
0.00

=

We need that instanceID will be included in this metrics or in some way we can specify that some metrics retrieve by generic aspect must be multidimension.
Last week we had a team session with R&D, so please assign them the case because they are aware of this matter.

Case "SD02796628 - AWS MP Monitoring CloudWatch Agent filesystem metrics with generic aspect" has been also raised for this purposal.

Many thanks.

[FSD1982]

We have exactly the same issue and also we have opened a support case and have a WebEx with RnD. They're evaluation the options... Also we need be able to graph FS utilization from CODA metrics, so they need to ensure to match metrics from CODA to OBM perfomance dashboards.

Yes, this is a very critical issue in our environment.

Please, can you update us with any information do you have? Thanks a lot!