Sorry for missing this, it went into my junk folder for some reason.
It seems to work for me, when using the above archive setting, I get an
rrd created with the following RRA:
RRA:AVERAGE:0.6:30:8640
What is your poll interval? That's the minimum timespan that can be
archived.
After removing the rrd file (or moving it out of the way), could you
post the output of the following command:
# rrdbot-create -nv
Thanks!
Stef
I created a new one for a handfull of counters, that i need to be
populated with 12 minute polls for 5 days with:
interval: 720
and
archive: 5/hour * 5 days
the resulting 'rrdbot-create -n' looks like:
# rrdtool create '/var/db/rrds/r1_ints.rrd' -b-1y -s720
DS:3_0:COUNTER:2160:0:U DS:4_0:COUNTER:2160:0:U DS:5_0:COUNTER:2160:0:U
DS:6_0:COUNTER:2160:0:U DS:3_0-d:COUNTER:2160:0:U
DS:4_0-d:COUNTER:2160:0:U DS:5_0-d:COUNTER:2160:0:U
DS:6_0-d:COUNTER:2160:0:U DS:3_1:COUNTER:2160:0:U DS:3_1
-d:COUNTER:2160:0:U DS:4_1:COUNTER:2160:0:U DS:4_1-d:COUNTER:2160:0:U
DS:5_1:COUNTER:2160:0:U DS:5_1-d:COUNTER:2160:0:U
DS:6_1:COUNTER:2160:0:U DS:6_1-d:COUNTER:2160:0:U RRA:AVERAGE:0.6:1:600
which looks OK in itself, but when i run an:
rrdtool fetch /var/db/rrds/r1_ints.rrd AVERAGE|wc -l
i get only 122 lines with timestamps returned.
which translates to 24h worth of datasets only
It seems i am missing something about how the rrdtool works, and that
this is not a rrdbot issue at all.
and i also spent sometime now with the rrdtool docu, but seem to miss
the casus-knacktus for this issue :-/
appreciate any help/pointers!!
TIA
-k
This defaults to the current day. See:
http://oss.oetiker.ch/rrdtool/doc/rrdfetch.en.html
Cheers,
Stef
The released version of rrdbot doesn't run on Solaris. However just put
together a machine with a OpenSolaris 2009-06 install, and so I spent
some time modifying it so that it does build and run.
Attached is the resulting patch. Please test this and let me know how it
goes. Once I hear back from you. I'll include this in the next release
of rrdbot.
Cheers,
Stef
kashmish wrote:
> but now to my problem:
> unfortunately my snmp-oid are dynamic, they can disappear over time.
> in that case the snmp query would return a 'Unknown variable name' for
> some of the individual items within the bundled request,
Does using 'table queries' solve your dynamic oid problem? See the TABLE
QUERIES section here:
http://memberwebs.com/stef/software/rrdbot/rrdbot.conf.5.html
> nd it seems that if a bundled rrdbot collection request fails even for a single oid within the bundled request,
> that all other values from that bundle also seem to get dropped?!
Is rrdbot dropping the 'other values' or is that your SNMP agent?
Cheers,
Stef
Stef Walter wrote:
> I hope it's okay if I CC this the rrdbot mailing list.
>
> kashmish wrote:
>
>> but now to my problem:
>> unfortunately my snmp-oid are dynamic, they can disappear over time.
>> in that case the snmp query would return a 'Unknown variable name' for
>> some of the individual items within the bundled request,
>>
>
> Does using 'table queries' solve your dynamic oid problem? See the TABLE
> QUERIES section here:
>
> http://memberwebs.com/stef/software/rrdbot/rrdbot.conf.5.html
>
unfortunately not, these are dynamic id, they are being re-created e.g.
each time the device on the other end would reboot powercycle or such,
no usefull and easy to identify label on the other end :-/
>> and it seems that if a bundled rrdbot collection request fails even for a single oid within the bundled request,
>> that all other values from that bundle also seem to get dropped?!
>> Is rrdbot dropping the 'other values' or is that your SNMP agent?
>>
>>
good question, it's a cisco router, would they behave like that? have
you ever observed that?
i will certainly dig into that on my end...
thx for the quickest response!!
-k
Use snmp2:// urls instead of snmp:// this will use snmp protocol two,
which supports requests that partially succeed/fail.
Cheers,
Stef