This is a discussion on [rrd-users] Re: seek error in RRA within the RRD Users forums, part of the Networking and Network Related category; Hi Mark, Let me tell you the background for this first. I produce a set of graphs for a specific ...
|
|||||||
| FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
|
|||
|
Hi Mark,
Let me tell you the background for this first. I produce a set of graphs for a specific interval. Then i compare graphs for different periods of time. In order to do this, I normalize the time period. I wrote a script to take a rrd and normalize it. Basically it copies the definition from an rrd, creates a new rrd with the new time period and updates the data with for a specific time range. Since I need to keep the definition of the rrd the same, and it's automated, it doesn't allow me to change the record length. The definition of the original rrd is exactly the same but for one difference, the beginning time. If you remove the -b and the argument, you can see that it will work fine. Also, if you remove some of the data sources, it also work fine. (even if you leave the -b untouched). This is why I beleive it to be a bug. As an addendum, I was using perl to parse the rrd, and I needed to see that the completed rrd contains the appropriate data sources. I use my ($names) = (RRDs::fetch("file","AVERAGE"))[2]; to find the names of them. Since i'm only interested in the names, and not the data, the range I'm fetching doesn't matter. I found that if I use: my ($names) = (RRDs::fetch("file", 'AVERAGE','-s',315532801,'-e',315532851))[2]; I could circumvent the seek error (it doesn't give one if I specify the range). It seems to be me it's a time issue more than anything. Btw, I'm normalizing to 1980, as I took over some work from someone else, and this is how they were doing it. db On Thu, Feb 24, 2005 at 08:05:27AM +0000, Mark Lawrence wrote: > Hi Daniel, > > > rrdtool create test.rrd -b 315532801 -s 30 DS:wrqm_s:GAUGE:300:U:U > > You specify here a step size of 30 seconds. > > > DS:r_s:GAUGE:300:U:U DS:w_s:GAUGE:300:U:U DS:rsec_s:GAUGE:300:U:U > > DS:wsec_s:GAUGE:300:U:U DS:rkB_s:GAUGE:300:U:U DS:wkB_s:GAUGE:300:U:U > > DS:avgrq_sz:GAUGE:300:U:U DS:avgqu_sz:GAUGE:300:U:U > > DS:await:GAUGE:300:U:U DS:svctm:GAUGE:300:U:U RRA:AVERAGE:0.5:1:1440 > > And you have here a single RRA with 1440 data points spaced 1 step size > apart (30 seconds). 1140 data points 30 seconds apart holds a maximum > history of 43200 seconds. > > By default rrdfetch or rrdgraph uses the previous 24 hours worth of data > which is 86400 seconds. So I suspect it is looking for data which is not > simply not there. > > I bet if you increased the number of points stored in the RRA to 2880 (or > maybe 2881?) it will work for you. > > Regards, > Mark. > -- > Mark Lawrence > > -- > Unsubscribe mailto:rrd-users-request@list.ee.ethz.ch?subject=unsubscribe > Help mailto:rrd-users-request@list.ee.ethz.ch?subject=help > Archive http://www.ee.ethz.ch/~slist/rrd-users > WebAdmin http://www.ee.ethz.ch/~slist/lsg2.cgi -- "Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety." -- Benjamin Franklin ----- End forwarded message ----- -- "Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety." -- Benjamin Franklin -- Unsubscribe mailto:rrd-users-request@list.ee.ethz.ch?subject=unsubscribe Help mailto:rrd-users-request@list.ee.ethz.ch?subject=help Archive http://www.ee.ethz.ch/~slist/rrd-users WebAdmin http://www.ee.ethz.ch/~slist/lsg2.cgi |
![]() |
| Thread Tools | |
| Display Modes | |
|
|