From: shanman@shanman.net
Assuming that you are doing 5 minute polls, you could add an average CF
RRA:
rra TotMonthAvg AVERAGE:0.5:1:8064
and maybe a max RRA:
rra TotMonthMax MAX:0.5:2:4032
Then you need to add them to the targetType dicts that you will be using.
The default target type lists:
rra = "5minAve, 30minAve, 2hrAve, 2hrMax, 1dayAve, 1dayMax"
So, you need to overwrite the 5MinAve with the TotMonthAvg. And the
30minAve as well since you are basically covering the same period. So:
rra = "TotMonthAvg, TotMonthMax, 2hrAve, 2hrMax, 1dayAve, 1dayMax"
This will grow your rrd file quite dramatically. So, try it on a test
datasource first. I haven't had good success with resizing current rrd
files with new rra's. So, I recommend you archive your old rrd's (if you
have a lot of data in them), and let cricket create new ones.
Shannon
On Wed, 3 Nov 1999, Mark Skinner wrote:
> From: "Mark Skinner" <mskinner@verio.net>
>
> Has anyone added or configured RRD to save _All_ the 5 min samples?
> I am hoping to save all the samples for a month to calculate 95% percentile
> values.
> also, it would be nice to calculate from the beginning of the month the
> current 95% percentile and put that on the graph as a bright red line.
>
> thanks
>
> Mark
>
> -----Original Message-----
> From: Hugo Gayosso <qztf7k@powertrain.mpg.gm.com>
> To: shane@bratnet.net <shane@bratnet.net>
> Cc: cricket-users@onelist.com <cricket-users@onelist.com>
> Date: Wednesday, November 03, 1999 5:58 AM
> Subject: RE: [cricket-users] Monitoring Load Average
>
>
> >From: Hugo Gayosso <qztf7k@powertrain.mpg.gm.com>
> >
> >-----BEGIN PGP SIGNED MESSAGE-----
> >Hash: SHA1
> >
> >I am monitoring the load average for some HP-UX machines, and I only have
> that
> >problem of the "mili" in one server, the rest are fine, here is my config
> file:
> >
> >
> >~/cricket-config/servers/Defaults
> >
> ># server-load sub-tree
> >#
> ># This is were we collect data from SNMP agents on HP-UX machines
> >
> >Target --default--
> > server-load = %auto-target-name%
> > snmp-host = %server-load%
> > # you could set a router-specific community string here:
> > # snmp-community = not-public
> > #
> > # Provide no default type, since each router will be different,
> > # and if we want to specify each model explicitly, or
> > # get a warning.
> > target-type = undef
> >
> ># OID's we'll be using in this tree
> >OID cpu1min 1.3.6.1.4.1.11.2.3.1.1.3.0
> >OID cpu5min 1.3.6.1.4.1.11.2.3.1.1.4.0
> >OID cpu15min 1.3.6.1.4.1.11.2.3.1.1.5.0
> >OID freemem 1.3.6.1.4.1.11.2.3.1.1.7.0
> >
> >datasource --default--
> > rrd-ds-type = GAUGE
> > rrd-heartbeat = 1800
> > rrd-min = undef
> > rrd-max = undef
> >
> >datasource cpu1min ds-source =
> snmp://%snmp%/cpu1min
> >datasource cpu5min ds-source =
> snmp://%snmp%/cpu5min
> >datasource freemem ds-source =
> snmp://%snmp%/freemem
> >datasource freeswap ds-source =
> snmp://%snmp%/freeswap
> >
> >targetType HP-UX
> > ds = "cpu1min, cpu5min, freemem"
> > view = "cpu: cpu1min cpu5min, memory: freemem"
> >
> >
> >html short-desc "HP-UX based machine"
> >
> >graph cpu1min
> > units = "loads"
> > color = dark-green
> > draw-as = AREA
> > y-axis = "Machine Load"
> > legend = "1 minute Load"
> > scale = "100,/"
> >
> >graph cpu5min
> > units = "loads"
> > color = blue
> > y-axis = "Machine Load"
> > legend = "5 minute Load"
> > scale = "100,/"
> >
> >graph freemem
> > color = blue
> > draw-as = LINE2
> > y-axis = "Bytes"
> > legend = "Free Memory"
> > bytes = true
> > scale = "1024,*"
> >
> >
> >
> >
> >
> >
> >P.S.
> > If you can provide me with an example to monitor Disk space, I
> would
> >appreciate it :)
> >
> > I have been busy to do some research about this using SNMP. Right
> now I
> >use Netsaint for that, and it uses a Perl daemon running in the server, but
> I
> >would like to do it via SNMP so I don't run two daemons (SNMP and Perl).
> >
> >Greetings,
> >
> >
> >
> >
> >On 03-Nov-99 shane@bratnet.net wrote:
> >> From: shane@bratnet.net
> >>
> >> Does anyone have the tricks needed to make rrdtool graph
> >> loadaverages corectly, as opposed to scalling them to 1 - 100mili etc.
> >>
> >> I have seen someone on this list ask this before, but I couldn't find the
> >> answer posted to the list. I am trying to make a set of config files to
> >> graph usage of my server for disk/memory/load average using UCD-SNMP, to
> >> monitor
> >> about 50 servers. I have all working except Getting the load average to
> graph
> >> corectly.
> >>
> >>
> >> Has anyone else set this up? Care to share your configs with me or the
> list?
> >>
> >> shane@bratnet.net
> >>
> >>
> >- --
> >Hugo Gayosso | Free Software implies
> >Controls Process Support |
> >Electronics Integration and Software | "Free speech"
> >GM Powertrain |
> >Brighton, MI | NOT
> >USA |
> > | "Free beer" !!!
> >E-Mail: qztf7k@powertrain.mpg.gm.com |
> > Phone: (810) 220-2635 | http://www.gnu.org
> > Fax: (810) 220-2663 |
> >
> >
> >
> >-----BEGIN PGP SIGNATURE-----
> >Version: GnuPG v1.0.0 (GNU/Linux)
> >Comment: For info see http://www.gnupg.org
> >
> >iD8DBQE4ID2ix2JZtTN6co8RAqsRAJ9p+ORSzWxTWk898jeM2VIiiTmZUQCfVoCh
> >5JcYOaPioLUJFhBUPnBZfJU=
> >=Wztg
> >-----END PGP SIGNATURE-----
> >
> >>
>
> >
--------------------------- ONElist Sponsor ----------------------------
Independent contractors: Find your next project gig through JobSwarm!
You can even make $$$ by referring friends.
<a href=" http://clickme.onelist.com/ad/jobswarm2 ">Click Here</a>
------------------------------------------------------------------------
This archive was generated by hypermail 2b29 : Mon Mar 06 2000 - 19:01:05 PST