Issue#1
My YACE config looks like below but sometimes cpu utilization or some other
metric reports value as Not a number(NaN) but I dont understand why because
underlying cloudwatch data points are correct. what is wrong here?
apiVersion: v1alpha1
discovery:
jobs:
- type: AWS/RDS
regions:
- us-east-1
- us-west-2
roles:
- roleArn: "arn:aws:iam::xxxxxxxxxx:role/yyyyy"
- roleArn: "arn:aws:iam:: xxxxxxxxxx :role/ yyyyy "
- roleArn: "arn:aws:iam:: xxxxxxxxxx :role/ yyyyy "
- roleArn: "arn:aws:iam:: xxxxxxxxxx :role/ yyyyy "
* period: 60 length: 60*
metrics:
- name: CPUUtilization
statistics: [Average]
- name: DatabaseConnections
statistics: [Average]
- name: FreeableMemory
statistics: [Average]
- name: FreeStorageSpace
statistics: [Average]
- name: ReadThroughput
statistics: [Average]
- name: WriteThroughput
statistics: [Average]
- name: ReadLatency
statistics: [Average]
- name: WriteLatency
statistics: [Average]
- name: ReadIOPS
Issue#2
Now exporter to cloudwatch scrape interval is 60 seconds, prometheus to
exporter scrape interval is 20 seconds and evaluation interval in
prometheus is 20 seconds.
Now my following rules are always stuck in peding state, how do I adjust my
configuration to make things work?
(aws_rds_write_latency_average{account_id!="",dimension_DBInstanceIdentifier!=""})
* 1000 > 20
<https://prometheus-cieops-monitoring.us-east-1.prod.cieops.aws.athenahealth.com/graph?g0.expr=%28aws_rds_write_latency_average%7Baccount_id%21%3D%22973732892259%22%2Cdimension_DBInstanceIdentifier%21%3D%22%22%7D%29+%2A+1000+%3E+20&g0.tab=1>
avg_over_time(aws_rds_write_latency_average{account_id!="973732892259",dimension_DBInstanceIdentifier!=""}[10m])
* 1000 > 10
<https://prometheus-cieops-monitoring.us-east-1.prod.cieops.aws.athenahealth.com/graph?g0.expr=avg_over_time%28aws_rds_write_latency_average%7Baccount_id%21%3D%22973732892259%22%2Cdimension_DBInstanceIdentifier%21%3D%22%22%7D%5B10m%5D%29+%2A+1000+%3E+10&g0.tab=1>
isue#3
sometimes when I manually verify the values between cloudwatch and
prometheus exporter, I see some fluctuations in cloudwatch data points
whereas prometheus is not reflecting that.
For example - every few minutes cloudwatch reports write latency as 80 ms
and then goes down to 10ms , stays there for 3-4 minutes and then goes up
whereas in grafana dashboard this fluction is not seen and it always shows
60-80 milliseconds for the entire hour.
Please help me to fix my configuration.
Regards,
Niranjan
--
You received this message because you are subscribed to the Google Groups
"Prometheus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion visit
https://groups.google.com/d/msgid/prometheus-users/630e054e-c2b1-49d6-9a33-10a3f57e93afn%40googlegroups.com.
