Jean-Michel Vourgère | 27 Aug 14:39 2015
Picon

Weird resolution

Hi

While investigating an issue with PREV(), I found myself unable to
explain the resolution in that case:
- An rrd file "esel.rrd" is created by "create.sh" and filled by "fill.py".

Data is available:
- Every 2 minutes for 12 hours
- Every 8 minutes for 48 hours
- ...

When I graph starting at "end-12h" or "end-11h", I expected a 2 minute
resolution. But what I get is more like a 8 minutes one.
See attached "graph.sh" and the resulting png.
Width is 720 pixels wide: One pixel per minute on a 12 hours period.
The resolution is ok for 10 hours (smooth curve), but 11 hours and 12
hours yield a stair effect.

Is that a bug?
Attachment (create.sh): application/x-shellscript, 231 bytes
Attachment (esel.rrd): application/octet-stream, 16 KiB
Attachment (fill.py): text/x-python, 343 bytes
Attachment (graph.sh): application/x-shellscript, 782 bytes
_______________________________________________
rrd-developers mailing list
rrd-developers <at> lists.oetiker.ch
https://lists.oetiker.ch/cgi-bin/listinfo/rrd-developers
(Continue reading)

Jean-Michel Vourgère | 26 Aug 13:48 2015
Picon

lua5.0 support

Hi

I tried lua bindings with lua 5.3 and it looks ok. :)

However, the test at bindings/lua/test.lua.bottom is obsolete:
table.getn() doesn't exist anymore since lua 5.2. This need to be
replaced by the operator #.

I was wondering whether we should drop support for lua 5.0 altogether?
Supporting versions 5.1 - released in 2006 - and above seems good enough.

See release dates here:
http://www.lua.org/versions.html

I could work on a patch if agreed. configure.ac will be reduced.

--

-- 
Nirgal
Jean-Michel Vourgère | 9 Aug 19:59 2015
Picon

unit test failures on some arch: rpn2 and xport1

Hi,

3 attached files are the result of test-suite.log on 3 different
architectures.

1. xport1
---------

I tried building 1.5.4 on all architectures, with rpn2 disabled, and I
get a bunch of issues in xport1:
https://buildd.debian.org/status/package.php?p=rrdtool
:(

Looking at hurd-i386 and armel logs, we can see there still are some
issues with rounding. Both arch fails, but not at in the same way.
Should we go DIFF8 instead of DIFF9 ? :/

This is just a rounding issue.

2. rpn2
-------

I disabled the test because it doesn't work on my amd64:
week result is 3600 where it should be 720.
However, this test works on hurd-i386 and armel. Oo

There's something very wrong here. I did not look into it yet.
Attachment (test-suite-amd64.log): text/x-log, 5796 bytes
(Continue reading)

Jean-Michel Vourgère | 26 Jul 01:37 2015
Picon

Hurd build failure

Hi

I'm having a look at the history of problems caused by usage of
PATH_MAX. There has been a few proposals since August 2013, but they
were based on a Debian / GNU Linux version that already contained a
patch for hurd, which was introduced back in 2009, and never made it to
the official repository.

Patches by Svante Signell only are about rrd_daemon.c and rrd_client.c.
So, I'm pretty sure they assume the other occurrences have been fixed.

Attached is the original patch by Marc Dequènes for rrd_graph.c,
rrd_graph.h and rrd_tool.c, refreshed against branch 1.5.

In Debian and derivatives, it's been applied to every version since
2009. I reviewed it and it still seems ok.
I think it's just missing a conditional free(im->graphfile) in im_free().

Before creating a pull request, I'd like you opinion about the #if usage:
On one hand, it's nice to have it, since we avoid a malloc and use the heap.
On the other hand, it makes the code more complex, and filename
allocation during graphical operation probably doesn't use a lot of
ressources compared to cairo ploting, so it doesn't seem worth the trouble.

I slightly prefer version that works everywhere, and would like to
remove the static length usage, so that the code is more simple. How
does that sound?
Shall I make a request against master or against the 1.5 branch?

I saw some questions in the list about whether hurd is broken for not
(Continue reading)

Tippana Santosh | 21 Jun 06:08 2015
Picon

Rrd developer

_______________________________________________
rrd-developers mailing list
rrd-developers <at> lists.oetiker.ch
https://lists.oetiker.ch/cgi-bin/listinfo/rrd-developers
santosh | 9 Jun 16:17 2015
Picon

RRD FILE Values generation Error


Problem:

            while am trying Graph generation for 1 hour values using JRobin.
RRD updated values and actual RRD file values are different, but it should
be same. Could you please tell me what wrong with that. I posted sample
results below.

                    RRD FILE ERROR

CreateTime: 1433850600 (CurrentTime)
TimeStamp : 300
HeartBeat : 600
DataSource Type : GAUGE
RRD Archive :  AVERAGE

UpdateTimeStamps  RRDFileTimestamps  updateValues  RRD actualValues

                 1433850600 :                : NAN
1433850925   :   1433850900 :        3943    : 3943.0   
1433851225   :   1433851200 :        1243    : 1468.0
1433851525   :   1433851500 :        8983    : 8338.0
1433851825   :   1433851800 :        1690    : 2297.75
1433852125   :   1433852100 :        4397    : 4171.416666666667
1433852425   :   1433852400 :        7964    : 7666.75
1433852725   :   1433852700 :        8364    : 8330.666666666666
1433853025   :   1433853000 :        6482    : 6638.833333333333
1433853325   :   1433853300 :        6544    : 6538.833333333333
1433853625   :   1433853600 :        7600    : 7512.0
1433853925   :   1433853900 :        2079    : 2539.083333333333
1433854225   :   1433854200 :        8194    : 7684.416666666667

--
View this message in context: http://rrd-mailinglists.937164.n2.nabble.com/RRD-FILE-Values-generation-Error-tp7582996.html
Sent from the RRDtool Developers Mailinglist mailing list archive at Nabble.com.
Daniel Letai | 27 May 15:41 2015
Picon

rpmbuild rrdtool-1.5.3 missing source1, patch1

When doing a simple
rpmbuild -ta rrdtool-1.5.3.tar.gz
I get the following error:
error: File php4-r1190.tar.gz: No such file or directory

I have located this file in the fedora pkgs repo:
http://pkgs.fedoraproject.org/repo/pkgs/rrdtool/php4-r1190.tar.gz/28074a9c368af013462631959ab90558/php4-r1190.tar.gz

But the next error is trickier:
error: Bad source: 
rrdtool-1.3.0-beta4-fix-rrd_update-in-php-bindings.patch: No such file 
or directory

The only mention of this patch I could locate is in a rhel mailing list:
https://www.redhat.com/archives/fedora-extras-commits/2008-April/msg01547.html

So I'd like to know whether that message is the entire patch, and if 
it's possible to include both patch1 and source1 in the tar file, or at 
least in the svn/git repos.

If the files are already there, kindly send a link, I couldn't find them.

Thanks,
-Dani_L.
Bruce Morgan | 23 Apr 09:42 2015
Picon

RRDGraph Error in 1.5.0 and 1.5.1

Using RRDtool version 1.4.7 I get a graph with the following command
(actually I use RRDs but no matter):
rrdtool graph /tmp/vic-crlt-pe1:25900_day.png --start -24h --end 1429773557
--width 820 --height 350 --title "vic-crlt-pe1 xe-0/0/3.201"  -a PNG
DEF:687_xe_0_0_3_201=/srv/netdata/scu/data/vic-crlt-pe1/rrd/687/687-xe-0_0_3_201.rrd:ds0:AVERAGE
DEF:o687_xe_0_0_3_201=/srv/netdata/scu/data/vic-crlt-pe1/rrd/687/687-xe-0_0_3_201.rrd:ds1:AVERAGE
CDEF:687_xe_0_0_3_201b=687_xe_0_0_3_201,8,*
CDEF:o687_xe_0_0_3_201b=o687_xe_0_0_3_201,-8,*
CDEF:o687_xe_0_0_3_201bits=o687_xe_0_0_3_201,8,*
CDEF:d_687_xe_0_0_3_201b=687_xe_0_0_3_201b,o687_xe_0_0_3_201b,+
LINE:687_xe_0_0_3_201b#000000: HRULE:0#000000:
LINE:o687_xe_0_0_3_201b#000000:
VDEF:687_xe_0_0_3_20195=687_xe_0_0_3_201b,95,PERCENT
VDEF:687_xe_0_0_3_201av=687_xe_0_0_3_201b,AVERAGE
VDEF:687_xe_0_0_3_201max=687_xe_0_0_3_201b,MAXIMUM
VDEF:687_xe_0_0_3_201c=687_xe_0_0_3_201b,LAST
VDEF:o687_xe_0_0_3_20195=o687_xe_0_0_3_201bits,95,PERCENT
VDEF:o687_xe_0_0_3_201av=o687_xe_0_0_3_201bits,AVERAGE
VDEF:o687_xe_0_0_3_201max=o687_xe_0_0_3_201bits,MAXIMUM
VDEF:o687_xe_0_0_3_201c=o687_xe_0_0_3_201bits,LAST
PRINT:687_xe_0_0_3_20195:%6.3lf PRINT:o687_xe_0_0_3_20195:%6.3lf
PRINT:687_xe_0_0_3_201av:%6.3lf PRINT:o687_xe_0_0_3_201av:%6.3lf
PRINT:687_xe_0_0_3_201max:%6.3lf PRINT:o687_xe_0_0_3_201max:%6.3lf
PRINT:687_xe_0_0_3_201c:%6.3lf PRINT:o687_xe_0_0_3_201c:%6.3lf
901x420
  -nan
  -nan
  -nan
  -nan
  -nan
  -nan
  -nan
  -nan

However in 1.5 I get the following:
rrdtool graph /tmp/vic-crlt-pe1:25900_day.png --start -24h --end 1429773557
--width 820 --height 350 --title "vic-crlt-pe1 xe-0/0/3.201"  -a PNG
DEF:687_xe_0_0_3_201=/srv/netdata/scu/data/vic-crlt-pe1/rrd/687/687-xe-0_0_3_201.rrd:ds0:AVERAGE
DEF:o687_xe_0_0_3_201=/srv/netdata/scu/data/vic-crlt-pe1/rrd/687/687-xe-0_0_3_201.rrd:ds1:AVERAGE
CDEF:687_xe_0_0_3_201b=687_xe_0_0_3_201,8,*
CDEF:o687_xe_0_0_3_201b=o687_xe_0_0_3_201,-8,*
CDEF:o687_xe_0_0_3_201bits=o687_xe_0_0_3_201,8,*
CDEF:d_687_xe_0_0_3_201b=687_xe_0_0_3_201b,o687_xe_0_0_3_201b,+
LINE:687_xe_0_0_3_201b#000000: HRULE:0#000000:
LINE:o687_xe_0_0_3_201b#000000:
VDEF:687_xe_0_0_3_20195=687_xe_0_0_3_201b,95,PERCENT
VDEF:687_xe_0_0_3_201av=687_xe_0_0_3_201b,AVERAGE
VDEF:687_xe_0_0_3_201max=687_xe_0_0_3_201b,MAXIMUM
VDEF:687_xe_0_0_3_201c=687_xe_0_0_3_201b,LAST
VDEF:o687_xe_0_0_3_20195=o687_xe_0_0_3_201bits,95,PERCENT
VDEF:o687_xe_0_0_3_201av=o687_xe_0_0_3_201bits,AVERAGE
VDEF:o687_xe_0_0_3_201max=o687_xe_0_0_3_201bits,MAXIMUM
VDEF:o687_xe_0_0_3_201c=o687_xe_0_0_3_201bits,LAST
PRINT:687_xe_0_0_3_20195:%6.3lf PRINT:o687_xe_0_0_3_20195:%6.3lf
PRINT:687_xe_0_0_3_201av:%6.3lf PRINT:o687_xe_0_0_3_201av:%6.3lf
PRINT:687_xe_0_0_3_201max:%6.3lf PRINT:o687_xe_0_0_3_201max:%6.3lf
PRINT:687_xe_0_0_3_201c:%6.3lf PRINT:o687_xe_0_0_3_201c:%6.3lf
*Unable to create graph: Not a valid vname: o687_xe_0_0_3_201bit in line
o687_xe_0_0_3_201bits,95,PERCENT *

So the error message shows a truncated vname 

In rrd_graph_helper.c:

   strncpy(gdp->ds_nam,gdp->rpn,DS_NAM_SIZE);
    *c=',';
    /* trying to find the vidx for that name */
    gdp->vidx = find_var(im, gdp->ds_nam);
    if (gdp->vidx<0) { *c=',';
      rrd_set_error("Not a valid vname: %s in line %s", gdp->ds_nam,
gdp->rpn);
      return 1;}

So it seems that the lookahead is getting the comma earlier than it should.
Anyone able to sort this?

regards

Bruce

--
View this message in context: http://rrd-mailinglists.937164.n2.nabble.com/RRDGraph-Error-in-1-5-0-and-1-5-1-tp7582906.html
Sent from the RRDtool Developers Mailinglist mailing list archive at Nabble.com.
Tobias Oetiker | 22 Apr 14:04 2015
Picon

[rrd] rrdtool 1.5.1 released

For all those who never install a x.0 version: RRDtool 1.5.1 is out.
Get a copy from http://www.rrdtool.org

A big thank you to the sponsors who are helping to fund my work on
rrdtool: http://www.wuerth-phoenix.com/ https://www.open.ch/
https://www.psw.net/, http://www.irisns.com/ and http://beslist.be

RRDtool 1.5.1 - 2015-04-22
==========================
Bug Fixes
--------
* parse numbers up to 40 characters long ..
* fix install rules for Python and Lua
* include missing VERSION and LICENSE file
* unlink before rename in rrd_create when running on WIN32
--

-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland
www.oetiker.ch tobi <at> oetiker.ch +41 62 775 9902
Tobias Oetiker | 16 Apr 23:48 2015
Picon

rrdtool version 1.5.0 is out

Finally! RRDtool 1.5.0 is out.
Get a copy from http://www.rrdtool.org

A big thank you to the sponsors who are helping to fund my work on
rrdtool: http://www.wuerth-phoenix.com/ https://www.open.ch/
https://www.psw.net/ and http://www.irisns.com/

New Features
------------
* automatic x-axis labels that work from 1s to 30y on a single chart
* librados integration
* new datasource types: DCOUNTER and DDERIVE (they work the same as
  the original DS, except that they can deal with floatingpoint
  numbers).
* compile without graphics libraries: ./configure --disable-rrd_graph
* updated windows port (see WIN32-BUILD-TIPS.txt)
* single step RRAs for MIN,MAX,LAST are generated virtually from
  an AVERAGE RRA
* ignore updates in the past with rrdtool update --skip-past-updates
* ignore a LINE when scaling a chart using the skipscale option
* detect 32bit timeoverflows
* massive performance boost for charts with more than 100 DEF line
* by switching form a linear search to a HASH when searching for data
* improved cross compilation support
* .Net bindings
* allow rrdtool graph to silently skip non-existing source files
  using the --use-nan-for-all-missing-data option
* restore from a pipe (rrdtool restore - y.rrd)
* in rrdtool create, row count and step can be defined in absolute time
* all new "create on steroids" can pull both data and configuration
* from existing rrd files
* use rrdtool graph to chart arbitrary data via a callback function
  for data fetching support is integrated in the perl bindings.
* re-written parser for rrdtool graph commands. It now follows a
  simple key  value pattern, compatible with the previous syntax.
* MEDIAN op for CDEF expressions
* DEPTH,INDEX,COPY,ROL ops for CDEF (as seen in PostScript)
* gradient AREA backgrounds
* no more locale magic while reading numeric data.

Bugfixes
--------
all the bugs fixed in 1.4.x during 1.5 development

--

-- 
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland
www.oetiker.ch tobi <at> oetiker.ch +41 62 775 9902
Gaby | 14 Mar 17:00 2015
Picon

RRDcached performance issues (from Users mailinglist)

Dear developers, could you take a look at my post in the Users forum? 

http://rrd-mailinglists.937164.n2.nabble.com/RRDcached-performance-issues-tp7582783.html;cid=1426345040002-291

It has no replies, and I'm still very interested in shedding some light on
it.

Gaby

--
View this message in context: http://rrd-mailinglists.937164.n2.nabble.com/RRDcached-performance-issues-from-Users-mailinglist-tp7582860.html
Sent from the RRDtool Developers Mailinglist mailing list archive at Nabble.com.

Gmane