I have sample with five min interval data is like this
time host job# mem_used (bytes)
00:05 host1 "123" 1,085,030,400
00:05 host2 "123" 1,085,030,500
00:05 host3 "124" 1,085,030,500
00:05 host4 "123" 1,085,030,400
00:05 host5 "125" 10,085,030,500
00:05 host6 "124" 2,085,030,500
what I need it range of uniq job running with max memory
0 to 8 gig ram -> 2 job
9 to 16gig ram -> 1 job (host 6 is using 10gig uniq count is 1)
job 123 & 124 are using max 0-8gig
job 125 is using max 9-16 gig
I think timelion can achieve this but I am not getting any success
> .es(index=myindex,interval=5m, timefield='@timestamp', metric=cardinality:job.keyword)
> .lines(width=1)
> .color(yellow)
> .label("0-8gb #Jobs"),
> .es(index=myindex,interval=5m, timefield='@timestamp', metric='max:mem_used').divide(1073741824).range(0,8)
> .label("Max mem_used - GB")
here whatever range I select it gives me same result.
What am I missing?