Average CPU usage for last 30 days, data granularity = "hourly"

[#1]

Hi,

we are trying to export csv/Json for average CPU usage for last 30 days (with data granularity "Hourly") using Cloudera manager and used following query:-

SELECT cpu_percent_across_hosts WHERE entityName = "1" AND category = CLUSTER

Queries:-

1. is this correct query/parameter to pull average cpu usage for cluster?  if not, please suggest on how can we get this data.

2. If answer to question 1 is yes, challenge is, If we take data for last 30 days, we do not get granularity lesser than 6 hours but if we select last 7 days, we get data grouped by 1 hour (which we need). can we do something about this?

thanks, 

Gaurav

@admin 

@support 

[#2]

@GauravCha 

Yes your query to get the cluster CPU usage is correct. 

Regarding the Data Granularity, in the chart builder you can select Show Additional Options next to 30d in the time frame. In the Additional Options you can select the granularity to be Hourly.

Kindly let me know if this is what you are looking for.

[#3]

@tjangid 

I am checking the metrics for Memory available for yarn, for the last 2 months data I can see the data with data granularity of every 6 hours but the previous months data shows only weekly in the chart even if I set the data granularity as per 6 hours. I am confused why it is so. any clue ??