Andrius Narbutas | 1 Nov 20:54 2005
Picon

Bugs : digits in vname, locale, grapg size

I recently wrote to rrd-users mailing list, but didn't get any hints how 
to solve problems, so... Maybe this is correct place:

Upgraded to rrdtool 1.2.11 with hope, that i get my regional letters
shown correctly ;-) Got several problems :

a) fixed colons in code. This isn't a problem, this is described in
"major incompat." list. BUT. I haven't find anywhere mentioned, that
variables in 1.2.x can't be only digits. Instead of that, in 
documentation 
http://people.ee.ethz.ch/~oetiker/webtools/rrdtool/doc/rrdgraph_data.en.html 

there is : "Variable names (vname) must be made up strings of the 
following characters A-Z, a-z, 0-9, -,_ and a maximum length of 255 
characters".

I have used such variables in 1.0.x version without any problems (shell 
script that picks users from db, finds rrd and draws graphs, so the 
easiest way to implement this - start a counter and use it as variable 
;-) - but now my scripts stopped working. Take a look :

rrdtool graph img.png --start -86400 --imgformat PNG --step 300
DEF:1=db.rrd:Lbig:AVERAGE LINE1:1#61C323:label

This works in 1.0.x without problems, but in 1.2.x it draws only empty
horiz. line. No warnings, no errors. Just don't work as expected. It
seems that it picks "counter value" as data value, and, if graph is
drawn in --rigid mode with upper and lower limits set to big values, you
just get an line  <at> 0.

(Continue reading)

Alex van den Bogaerdt | 1 Nov 22:04 2005
Picon

Re: Bugs : digits in vname, locale, grapg size

On Tue, Nov 01, 2005 at 09:54:44PM +0200, Andrius Narbutas wrote:
> I recently wrote to rrd-users mailing list, but didn't get any hints how 
> to solve problems, so... Maybe this is correct place:

http://lists.ee.ethz.ch/rrd-users/msg10522.html

--

-- 
Alex van den Bogaerdt
http://www.vandenbogaerdt.nl/rrdtool/

--
Unsubscribe mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=help
Archive     http://lists.ee.ethz.ch/rrd-developers
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi

Andrius Narbutas | 1 Nov 23:47 2005
Picon

Re: Bugs : digits in vname, locale, grapg size

On 2005.11.01 23:04, Alex van den Bogaerdt wrote:

> http://lists.ee.ethz.ch/rrd-users/msg10522.html

Excuse me, i do not want to start a flame here, but... I'm wrong? Was
your e-mail helpful a bit? I just said that :
a) rrdtool documentation permits using digits as vnames
b) it WorksForMe in v 1.0.x.

No more, no less. Maybe it is error prone, maybe it is incorrect, but it
worked, and stopped working. And i just asked to update docs. Thanks for
help, but i didn't find anything usefull.

About encodings... I searched in devel lists about that problem... And
didn't find related docs. Maybe, except string in rrd_graph.c, which, as
i already said, isn't working. Maybe i didn't know how to search, maybe
i can't find what i need. But u didn't wrote too. So thanks again for
direction, but it didn't help me (as i already searched lists before
writing). Link to your answer is good, but i already have this e-mail in
my inbox ;-)

And last - i do not know why your mutt incorrectly wraps my emails. My
mutt wraps it without problems. Strange.

Sorry for offtopic. Can anybody give hint how to debug incorrect encoding?

--
Unsubscribe mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=help
Archive     http://lists.ee.ethz.ch/rrd-developers
(Continue reading)

Edmonds, Alan | 2 Nov 10:57 2005
Picon

rrdupdate error - Not a simple integer


I started seeing this error after upgrading from RRDtool 1.0.x to 1.2.

Perhaps I'm missing something with rrdtool.  I'm using rrdtool fronted 
by Cricket 1.0.5 (http://cricket.sourceforge.net/).

Quite simply, it appears you cannot use non-integers with rrdupdate with 
DERIVE and COUNTER types.  If a COUNTER is bits per second or units, it makes
sense that it would only be an integer.  But if it's something like megabits
per second or other non-integer values, it doesn't allow it.

I threw together a test script. It's the example from the rrdtutuorial man page
with the data values changed to floating point numbers.

	echo test 1 - GAUGE type

	rm -f test-gauge.rrd

	rrdtool create test-gauge.rrd  --start 920804400 DS:speed:GAUGE:600:U:U RRA:AVERAGE:0.5:1:24 RRA:AVERAGE:0.5:6:10
	rrdtool update test-gauge.rrd 920804700:1.2345 920805000:1.2357 920805300:1.2363
      rrdtool update test-gauge.rrd 920805600:1.2363 920805900:1.2363 920806200:1.2373
      rrdtool update test-gauge.rrd 920806500:1.2383 920806800:1.2393 920807100:1.2399
      rrdtool update test-gauge.rrd 920807400:1.2405 920807700:1.2411 920808000:1.2415
      rrdtool update test-gauge.rrd 920808300:1.2420 920808600:1.2422 920808900:1.2423

	echo test 2 - COUNTER type

	rm -f test-counter.rrd

	rrdtool create test-counter.rrd  --start 920804400 DS:speed:COUNTER:600:U:U RRA:AVERAGE:0.5:1:24 RRA:AVERAGE:0.5:6:10
(Continue reading)

Tobias Oetiker | 2 Nov 12:44 2005
Picon
Picon

Re: rrdupdate error - Not a simple integer

Hi Alan,

the problem is that rrdtool handles updates to DERIVE and COUNTER
by building the difference of two long long long long integers so
it this was changed to double then your request would be possible
... except that in this case we would be loosing precision at the
low end if you feed the system with counters that are at their max
which might be rather undesirable ...

cheers
tobi

>
> I started seeing this error after upgrading from RRDtool 1.0.x to 1.2.
>
> Perhaps I'm missing something with rrdtool.  I'm using rrdtool fronted
> by Cricket 1.0.5 (http://cricket.sourceforge.net/).
>
> Quite simply, it appears you cannot use non-integers with rrdupdate with
> DERIVE and COUNTER types.  If a COUNTER is bits per second or units, it makes
> sense that it would only be an integer.  But if it's something like megabits
> per second or other non-integer values, it doesn't allow it.
>
> I threw together a test script. It's the example from the rrdtutuorial man page
> with the data values changed to floating point numbers.
>
> 	echo test 1 - GAUGE type
>
> 	rm -f test-gauge.rrd
>
(Continue reading)

Tobias Oetiker | 2 Nov 13:16 2005
Picon
Picon

Re: rrdupdate error - Not a simple integer

Alan,

long long long long is just a placeholder for a long integer
number so it you convert a float/double to this then you loose
the decimal part ...

cheers
tobi

--

-- 
 ______    __   _
/_  __/_  / /  (_) Oetiker  <at>  ISG.EE, ETL F24.2, ETH, CH-8092 Zurich
 / // _ \/ _ \/ /  System Manager, Time Lord, Coder, Designer, Coach
/_/ \.__/_.__/_/   http://people.ee.ethz.ch/oetiker +41(0)44-632-5286

--
Unsubscribe mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=help
Archive     http://lists.ee.ethz.ch/rrd-developers
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi

Edmonds, Alan | 2 Nov 14:30 2005
Picon

Re: rrdupdate error - Not a simple integer

> From: Tobias Oetiker [mailto:oetiker <at> ee.ethz.ch]
> 
> Alan,
> 
> long long long long is just a placeholder for a long integer
> number so it you convert a float/double to this then you loose
> the decimal part ...

Sounds like I need to rollback to 1.0.x to keep 
the precision.  Have it been considered to include
a high precision math package into rrdtool
instead of rolling your own?  (rrd_diff)

Alan

--
Unsubscribe mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=help
Archive     http://lists.ee.ethz.ch/rrd-developers
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi

Tobias Oetiker | 2 Nov 14:33 2005
Picon
Picon

Re: rrdupdate error - Not a simple integer

Alan,

note that rrdtool 1.0 silently truncated the numbers ... I found
that this would get people to get the wrong idea hence I added the
error message ...

gruss
tobi
--

-- 
 ______    __   _
/_  __/_  / /  (_) Oetiker  <at>  ISG.EE, ETL F24.2, ETH, CH-8092 Zurich
 / // _ \/ _ \/ /  System Manager, Time Lord, Coder, Designer, Coach
/_/ \.__/_.__/_/   http://people.ee.ethz.ch/oetiker +41(0)44-632-5286

--
Unsubscribe mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-developers-request <at> list.ee.ethz.ch?subject=help
Archive     http://lists.ee.ethz.ch/rrd-developers
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi

D. Walsh | 3 Nov 03:09 2005

dump/restore

In my attempt to upgrade my rrdtool based apps I've run into a snafu  
regarding dump/restore.

Original version is 1.0.49, new is 1.2.11.

from reading the man pages and the rrdtool documentation, it appears  
that you are required to dump the old rrd files using the older  
rrdtools binary and restore using the newer rrdtool binary however,  
when I attempt to do so, it fails with the following error.
ERROR: Can only restore version >= 2 (Not 0001). Dump your old data  
using a current rrdtool dump.

Does this mean that the newer rrdtool will properly export the data  
from the older files???

-- Dale

--
Unsubscribe mailto:rrd-users-request <at> list.ee.ethz.ch?subject=unsubscribe
Help        mailto:rrd-users-request <at> list.ee.ethz.ch?subject=help
Archive     http://lists.ee.ethz.ch/rrd-users
WebAdmin    http://lists.ee.ethz.ch/lsg2.cgi

D. Walsh | 5 Nov 20:14 2005

cgilib-0.5.1.tar.gz

I understand this library is required and with all of the trouble  
required to install it I figured something should be done about it  
becuase it hasn't been worked on since 08/1999.
I took a few moments and fixed it, I've tested under Linux, Darwin,  
Mac OSX and it builds without any issues.

I've fixed the malloc.h issue, the problem with the cgitest.c and   
jumpto.c not compiling and an issue with string.h and have also built  
using gcc 4.0 without issue.

I've also added the ability to generate a shared library if you want  
one and the cgi-bin directory is definable during configure so you  
can point it to your favorite location.

To assist everyone one requiring cgilib in building rrdtool, IMHO it  
would be wise to place this in the rrdtool archives and amend the  
documentation to download and use this archive instead of the older  
cgilib-0.5.tar.gz archive.

I've attached it and hopefully it's not stripped off because upload  
access isn't available at the rrdtool site.

-- Dale

-- Attached file removed by Ecartis and put at URL below --
-- Type: application/x-gzip
-- Size: 296k (303116 bytes)
-- URL : http://lists.ee.ethz.ch/p/cgilib-0.5.1.tar.gz

-- Binary/unsupported file stripped by Ecartis --
(Continue reading)


Gmane