You can hide series with all nulls/zeroes in graph legend options in
grafana.
As for splitting by tags, I'd rather prefer allowing queries like "select
mean(*) ..." which would apply mean to each individual field, but aliasing
is tricky in this case.

On Tue, Jul 26, 2016 at 11:32 AM Jan Vokas <[email protected]> wrote:

> Let me explain what i mean:
>
> Telegraf sends data to influxdb via udp in this format (fe. disk):
> disk,datacenter=XXX,fstype=tmpfs,host=HOSTNAME,path=/run,rack=1a
> free=1495953408i,inodes_free=2051914i,inodes_total=2052602i,inodes_used=688i,total=1681494016i,used=185540608i,used_percent=11.034271084792252
> 1469520283896909617
> disk,datacenter=XXX,fstype=ext4,host=HOSTNAME,path=/,rack=1a
> free=96700481536i,inodes_free=6970042i,inodes_total=7282688i,inodes_used=312646i,total=117230362624i,used=20513103872i,used_percent=17.500619745226178
> 1469520283896966437
> (multiple values per line)
>
> It would be nice to have this format (separated values with tag path and
> tak valuename)
> disk,datacenter=XXX,fstype=tmpfs,host=HOSTNAME,path=/run,rack=1a,vname=free
> value=1495953408i 1469520283896909617
> disk,datacenter=XXX,fstype=tmpfs,host=HOSTNAME,path=/run,rack=1a,vname=inodes_total
> value=2052602i 1469520283896909617
> disk,datacenter=XXX,fstype=tmpfs,host=HOSTNAME,path=/run,rack=1a,vname=inodes_used
> value=688i 1469520283896909617
> and so on (one value per line)
>
> I "need" this for grafana graphs - because in default scenario, i'm not
> able to create graf with something like select * from disk and show all
> values in one graph. I must select valuename from disk metric for each
> valuename - some machines not sending all values, so there is hosts with
> empty values in graph.
>
> In the separated (and vname tagged) scenario - i will be able to create
> graphs like: select mean(value) from disk .. group by tag path, ... and all
> values sent by host will show up in one graph with one "select" statement,
> or i will be able to create select mean(value) from disk where vname =~
> ("/^inodes_.*$/)... for inodes graph and so on.
>
> It would be nice to have configuration option to set sending format of
> tags.
> Something global or per metric?
> SPLITVALUES={yes|no} (default no)
> VALUENAMETAG=vname (tagname for splitted scenario)
>
> Or is there another option which should i use to achieve same behaviour?
>
> ps.
> i have python script which fetch my old munin-nodes via telnet and send
> their metrics in this separated vname/value way and creating graphs for all
> hosts is very simple selects per graph. In the first version of my script -
> i send data to influx in the "grouped" format and creating dash/graphs was
> pain in the eye.
>
> Thanks
>
> --
> Remember to include the InfluxDB version number with all issue reports
> ---
> You received this message because you are subscribed to the Google Groups
> "InfluxDB" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at https://groups.google.com/group/influxdb.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/influxdb/e4a00138-b871-4d0f-95d5-053302bc555e%40googlegroups.com
> <https://groups.google.com/d/msgid/influxdb/e4a00138-b871-4d0f-95d5-053302bc555e%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
> For more options, visit https://groups.google.com/d/optout.
>

-- 
Remember to include the InfluxDB version number with all issue reports
--- 
You received this message because you are subscribed to the Google Groups 
"InfluxDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/influxdb.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/influxdb/CAD%2BeXGQB2FhWFD5QF922oqyLbWU9wNf8w5vF7uNxFEbcShqCew%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to