I am trying to emulate a bit of functionality of some of the better known SAAS metric visualization providers using InfluxDB and Grafana.
I am capturing process stats and inserting them into InfluxDB, I have pasted the Influx line format I am using and a simple aggregation that give me the rss of all the process over a period of time. Of course I do not want to graph all of the process, I only want to graph the top 5 or 10 process for each aggregation period.
I have done a little bit of google fu and I some people are using the TOP function and sub query to get the top N of something from the overall period, but what I want to get the the top N of something from each period.
My inclination is that I am going to have to use continuous queries to sample each collection period take the top N of the metric I care about and place that in a new measurement that I graph. I would have to create different measurements foe each thing I want to graph, rss, cputime, pcpu, âŠ
Can any suggest a way/ single query that I can use that will give me what I want.
All/Any help is much appreciated.
pslist,pid=579,user=root,comm=rpc.gssd cputime=11i,pcpu=0.0,pmem=0.0,rss=3992i,vsz=31388i
pslist,pid=584,user=root,comm=systemd-logind cputime=195i,pcpu=0.0,pmem=0.0,rss=3380i,vsz=22988i
pslist,pid=586,user=root,comm=accounts-daemon cputime=2187i,pcpu=0.0,pmem=0.0,rss=7352i,vsz=275284i
pslist,pid=643,user=root,comm=rngd cputime=1363i,pcpu=0.0,pmem=0.0,rss=140i,vsz=9512i
pslist,pid=688,user=root,comm=irqbalance cputime=530i,pcpu=0.0,pmem=0.0,rss=264i,vsz=19568i
pslist,pid=733,user=root,comm=agetty cputime=0i,pcpu=0.0,pmem=0.0,rss=1844i,vsz=9716i
pslist,pid=736,user=root,comm=agetty cputime=0i,pcpu=0.0,pmem=0.0,rss=1604i,vsz=9900i
pslist,pid=767,user=ntp,comm=ntpd cputime=1126i,pcpu=0.0,pmem=0.0,rss=5256i,vsz=106064i
pslist,pid=803,user=nslcd,comm=nslcd cputime=2060i,pcpu=0.0,pmem=0.0,rss=7348i,vsz=453680i
pslist,pid=809,user=root,comm=cron cputime=56i,pcpu=0.0,pmem=0.0,rss=2704i,vsz=22940i
pslist,pid=813,user=daemon,comm=atd cputime=0i,pcpu=0.0,pmem=0.0,rss=1932i,vsz=19716i
pslist,pid=960,user=root,comm=master cputime=85i,pcpu=0.0,pmem=0.0,rss=4484i,vsz=61228i
pslist,pid=975,user=postfix,comm=qmgr cputime=19i,pcpu=0.0,pmem=0.0,rss=4420i,vsz=65444i
pslist,pid=2140,user=root,comm=systemd-journal cputime=1299i,pcpu=0.0,pmem=1.8,rss=154380i,vsz=248928i
pslist,pid=4410,user=root,comm=kworker/1:2H cputime=1i,pcpu=0.0,pmem=0.0,rss=0i,vsz=0i
pslist,pid=4411,user=root,comm=kworker/2:2H cputime=2i,pcpu=0.0,pmem=0.0,rss=0i,vsz=0i
pslist,pid=4412,user=root,comm=kworker/0:2H cputime=3i,pcpu=0.0,pmem=0.0,rss=0i,vsz=0i
pslist,pid=4413,user=root,comm=kworker/3:2H cputime=3i,pcpu=0.0,pmem=0.0,rss=0i,vsz=0i
pslist,pid=6874,user=root,comm=sshd cputime=50i,pcpu=0.0,pmem=0.0,rss=6208i,vsz=61340i
pslist,pid=7991,user=nagios,comm=nrpe cputime=148i,pcpu=0.0,pmem=0.0,rss=2504i,vsz=19788i
pslist,pid=9008,user=statd,comm=rpc.statd cputime=1i,pcpu=0.0,pmem=0.0,rss=2892i,vsz=35368i
pslist,pid=9009,user=root,comm=rpcbind cputime=36i,pcpu=0.0,pmem=0.0,rss=3440i,vsz=47628i
SELECT mean(rss) AS rss FROM "30days"."pslist" WHERE time > now() - 5m AND "host" = 'xxxxxxxx' GROUP BY time(1m),"comm" SLIMIT 5
name: pslist
tags: comm=smartd
time rss
---- ---
1518792300000000000 3548
1518793200000000000 3548
1518794100000000000 3548
1518795000000000000 3548
name: pslist
tags: comm=sshd
time rss
---- ---
1518792300000000000 3288
1518793200000000000 3288
1518794100000000000 3288
1518795000000000000 3288
name: pslist
tags: comm=systemd
time rss
---- ---
1518792300000000000 4824.8
1518793200000000000 4824.8
1518794100000000000 4824.8
1518795000000000000 4824.354838709677
name: pslist
tags: comm=systemd-journal
time rss
---- ---
1518792300000000000 64906.4406779661
1518793200000000000 63610.2
1518794100000000000 63362.73333333333
1518795000000000000 64452
name: pslist
tags: comm=systemd-logind
time rss
---- ---
1518792300000000000 3532
1518793200000000000 3532
1518794100000000000 3532
1518795000000000000 3532
name: pslist
tags: comm=systemd-udevd
time rss
---- ---
1518792300000000000 3336
1518793200000000000 3336
1518794100000000000 3336
1518795000000000000 3336
