Mark Bednarczyk | 2 Aug 17:30 2004
Picon

RE: RRD/Format.pm support for s390/Linux

Oops, sorry. I got my timezones mixed up I put it down in my
calendar for 12:00pm my time.

Any time this afternoon or tomorrow.

Cheers,
mark...

> -----Original Message-----
> From: cricket-users-admin <at> lists.sourceforge.net
> [mailto:cricket-users-admin <at> lists.sourceforge.net]On Behalf Of
> shane <at> stixrud.org
> Sent: Tuesday, January 13, 2004 4:44 PM
> To: cricket-users <at> lists.sourceforge.net
> Subject: [cricket-users] RRD/Format.pm support for s390/Linux
>
>
> After searching cricket-users and google, I have
> found a number of work
> arounds for various platforms experiencing the
> "Current values not
> available: Architecture s390x-linux-thread-multi not
> supported yet"
> error in the summary window,  but not z390/linux (64Bit).  I
> was wondering if a workable solution would be to
> rename "sparc64-netbsd"
> to s390x-linux-thread-multi.
>
> Thanks.
>
(Continue reading)

Eugene Eug | 2 Aug 18:15 2004
Picon

bug in cricket diskbusy stats reporting

Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
Asif Saeed | 4 Aug 07:11 2004

SUNS OID

Hi all,

Anybody monitoring sun box , please send me OID/Script to monitor
SUN-box (disk usage, interface bandwaidh, processes) ,thanks

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
Gregory Bond | 4 Aug 07:56 2004
Picon

Integer / on-off data sources showing up as fractions


[I'm new to this list.  Looking at the mailing list archive, Tory Blue had a
similar question recently but there are no followups that I can see.]

I'm trying to graph yes-no variables (NMEA GPS synced yes/no, in this case.)
The number varies a bit atm while we try and find the best spot for the
antenna, but it's always either 0 or 1 (alas, mostly 0...)

I can see that the data going into RRD is integer.  The RRD is a GAUGE type.

But the graphs never show integer values!  The graph has never gone above 0.5!!
It jumps around a bit but it looks like there is some sort of multi-sample
smoothing going on, which is not at all what I expect.

I have other similar graphs that are integers (satellites in view, VPN links
active, etc) and they show similar behaviour.  

A sample of this graph (with not much data on it because I've re-created it 
several times trying to work out what's going on) can be found here:
	http://www.itga.com.au/~gnb/cricket-sample.png

Any clues???!!?

cricket <at> buzz$ ~/cricket/collector -logLevel debug /gps/status
[04-Aug-2004 15:48:06 ] Log level changed from warn to debug.
[04-Aug-2004 15:48:06 ] Starting collector: Cricket version 1.0.4 (2003-11-22)
[04-Aug-2004 15:48:06 ] Processing /gps/status...
[04-Aug-2004 15:48:06 ] target status
        archive-file = /home/cricket/cricket-config/../cricket-data//gps/status.archive
        auto-base = /home/cricket/cricket-config
        auto-root = /..
        auto-target-path = /gps
        copy-to = file:/home/cricket/cricket-config/../cricket-data//gps/status.archive
        data-file = /home/cricket/cricket-config/../cricket-data//gps/clockinfo
        datadir = /home/cricket/cricket-config/../cricket-data//gps
        directory-desc = GPS Statistics
        email-program = /usr/bin/mailx
        height-h = 255
        host = status
        inst = 
        long-desc = Sync Status
        order = 10
        persistent-alarms = false
        router = Router not specified in gps/Defaults
        rrd-datafile = /home/cricket/cricket-config/../cricket-data//gps/status.rrd
        rrd-poll-interval = 300
        short-desc = Sync Status
        show-path = no
        snmp = public <at> status:161:2.0:5:1.0:1
        snmp-backoff = 1.0
        snmp-community = public
        snmp-host = status
        snmp-port = 161
        snmp-retries = 5
        snmp-timeout = 2.0
        snmp-version = 1
        summary-loc = top
        target-type = nmea-status

[04-Aug-2004 15:48:06 ] Retrieving data for target status () (nmea-status)
[04-Aug-2004 15:48:06 ] ds0 is: /home/cricket/cricket-config/../cricket-data//gps/clockinfo:S:7:4:,
[04-Aug-2004 15:48:06 ] Reading data from
/home/cricket/cricket-config/../cricket-data//gps/clockinfo for status
[04-Aug-2004 15:48:06 ] Retrieved data for status (): 0
[04-Aug-2004 15:48:06 ] Appending to copy-to file /home/cricket/cricket-config/../cricket-data//gps/status.archive
[04-Aug-2004 15:48:06 ] Processing /gps/status...
[04-Aug-2004 15:48:06 ] Processed 1 targets in 0 seconds.

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
Marc Jimenez | 4 Aug 08:35 2004
Picon

Re: Integer / on-off data sources showing up as fractions

Hi Gregory,
	In the graph definition for your target, try including the line:
	"precision       =       integer"

	I believe this will have the effect you're looking for.

	-Marc

Marc Jimenez
Manager, Network Engineering and Security
Tufts University

On Wed, 4 Aug 2004, Gregory Bond wrote:

>
> [I'm new to this list.  Looking at the mailing list archive, Tory Blue had a
> similar question recently but there are no followups that I can see.]
>
> I'm trying to graph yes-no variables (NMEA GPS synced yes/no, in this case.)
> The number varies a bit atm while we try and find the best spot for the
> antenna, but it's always either 0 or 1 (alas, mostly 0...)
>
> I can see that the data going into RRD is integer.  The RRD is a GAUGE type.
>
> But the graphs never show integer values!  The graph has never gone above 0.5!!
> It jumps around a bit but it looks like there is some sort of multi-sample
> smoothing going on, which is not at all what I expect.
>
> I have other similar graphs that are integers (satellites in view, VPN links
> active, etc) and they show similar behaviour.
>
> A sample of this graph (with not much data on it because I've re-created it
> several times trying to work out what's going on) can be found here:
> 	http://www.itga.com.au/~gnb/cricket-sample.png
>
> Any clues???!!?
>
> cricket <at> buzz$ ~/cricket/collector -logLevel debug /gps/status
> [04-Aug-2004 15:48:06 ] Log level changed from warn to debug.
> [04-Aug-2004 15:48:06 ] Starting collector: Cricket version 1.0.4 (2003-11-22)
> [04-Aug-2004 15:48:06 ] Processing /gps/status...
> [04-Aug-2004 15:48:06 ] target status
>         archive-file = /home/cricket/cricket-config/../cricket-data//gps/status.archive
>         auto-base = /home/cricket/cricket-config
>         auto-root = /..
>         auto-target-path = /gps
>         copy-to = file:/home/cricket/cricket-config/../cricket-data//gps/status.archive
>         data-file = /home/cricket/cricket-config/../cricket-data//gps/clockinfo
>         datadir = /home/cricket/cricket-config/../cricket-data//gps
>         directory-desc = GPS Statistics
>         email-program = /usr/bin/mailx
>         height-h = 255
>         host = status
>         inst =
>         long-desc = Sync Status
>         order = 10
>         persistent-alarms = false
>         router = Router not specified in gps/Defaults
>         rrd-datafile = /home/cricket/cricket-config/../cricket-data//gps/status.rrd
>         rrd-poll-interval = 300
>         short-desc = Sync Status
>         show-path = no
>         snmp = public <at> status:161:2.0:5:1.0:1
>         snmp-backoff = 1.0
>         snmp-community = public
>         snmp-host = status
>         snmp-port = 161
>         snmp-retries = 5
>         snmp-timeout = 2.0
>         snmp-version = 1
>         summary-loc = top
>         target-type = nmea-status
>
> [04-Aug-2004 15:48:06 ] Retrieving data for target status () (nmea-status)
> [04-Aug-2004 15:48:06 ] ds0 is: /home/cricket/cricket-config/../cricket-data//gps/clockinfo:S:7:4:,
> [04-Aug-2004 15:48:06 ] Reading data from
/home/cricket/cricket-config/../cricket-data//gps/clockinfo for status
> [04-Aug-2004 15:48:06 ] Retrieved data for status (): 0
> [04-Aug-2004 15:48:06 ] Appending to copy-to file /home/cricket/cricket-config/../cricket-data//gps/status.archive
> [04-Aug-2004 15:48:06 ] Processing /gps/status...
> [04-Aug-2004 15:48:06 ] Processed 1 targets in 0 seconds.
>
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by OSTG. Have you noticed the changes on
> Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
> one more big change to announce. We are now OSTG- Open Source Technology
> Group. Come see the changes on the new OSTG site. www.ostg.com
> _______________________________________________
> cricket-users mailing list
> cricket-users <at> lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/cricket-users
>

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
Gregory Bond | 4 Aug 09:06 2004
Picon

Re: Integer / on-off data sources showing up as fractions

Thanks for the response.

mjimenez <at> net.tufts.edu said:
> 	"precision       =       integer" 

That changes the HTML summary info, but doesn't affect the graph.

Any more hints?

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
Marc Jimenez | 4 Aug 09:27 2004
Picon

Re: Integer / on-off data sources showing up as fractions

Hi Greg,
	I may be off-base here (I'm certainly not an expert), but I think
that what you're asking for may not be possible given the way RRDs
function. Since your device is sometimes a 1 and sometimes a 0, the RRDs
will eventually need to aggregrate these values. This is done by averaging
them, and therefore you wind up with fractional values.
	It seems likely to me that within the "daily" graphs where
presumably no summarization is occuring you could have strict integer
values for each sample period, but once you start the summarization
process, that's bound to change.

	Sorry that's not more helpful,

		Marc

Marc Jimenez
Manager, Network Engineering and Security
Tufts University

On Wed, 4 Aug 2004, Gregory Bond wrote:

> Thanks for the response.
>
> mjimenez <at> net.tufts.edu said:
> > 	"precision       =       integer"
>
> That changes the HTML summary info, but doesn't affect the graph.
>
> Any more hints?
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by OSTG. Have you noticed the changes on
> Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
> one more big change to announce. We are now OSTG- Open Source Technology
> Group. Come see the changes on the new OSTG site. www.ostg.com
> _______________________________________________
> cricket-users mailing list
> cricket-users <at> lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/cricket-users
>

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
Gregory Bond | 4 Aug 09:49 2004
Picon

Re: Integer / on-off data sources showing up as fractions


mjimenez <at> net.tufts.edu said:
> It seems likely to me that within the "daily" graphs where presumably
> no summarization is occuring you could have strict integer values for
> each sample period,

I understand weekly etc will have summaries that are averages, but this is the
daily graph I'm worried about at this point.  How can the "current" reading 
be 0.4 when all that are available is 0 or 1???

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
Clifton Royston | 4 Aug 19:16 2004

Re: Integer / on-off data sources showing up as fractions

On Wed, Aug 04, 2004 at 05:49:05PM +1000, Gregory Bond wrote:
> 
> mjimenez <at> net.tufts.edu said:
> > It seems likely to me that within the "daily" graphs where presumably
> > no summarization is occuring you could have strict integer values for
> > each sample period,
> 
> I understand weekly etc will have summaries that are averages, but this is the
> daily graph I'm worried about at this point.  How can the "current" reading 
> be 0.4 when all that are available is 0 or 1???

If there is one person in a room for half a day, and nobody in the room
for half a day, what is the average number of people in the room?  The
answer is 0.5; not 0 or 1.

If your value is 1 for part of a day, and 0 for part of the day,
presenting either 0 or 1 as an average would be wrong.

What you want is something other than an average.  RRD does averages.

  -- Clifton

--

-- 
          Clifton Royston  --  cliftonr <at> tikitechnologies.com 
         Tiki Technologies Lead Programmer/Software Architect
Did you ever fly a kite in bed?  Did you ever walk with ten cats on your head?
  Did you ever milk this kind of cow?  Well we can do it.  We know how.
If you never did, you should.  These things are fun, and fun is good.
                                                                 -- Dr. Seuss

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
Gregory Bond | 5 Aug 01:42 2004
Picon

Re: Integer / on-off data sources showing up as fractions


cliftonr <at> tikitechnologies.com said:
> If there is one person in a room for half a day, and nobody in the
> room for half a day, what is the average number of people in the room?
>  The answer is 0.5; not 0 or 1.

> If your value is 1 for part of a day, and 0 for part of the day,
> presenting either 0 or 1 as an average would be wrong.

But it's not an average!  It's the current value that is wrong.

A person is either in the room or not in the room.  Schrodinger aside, there 
cannot be 0.4 people in a room at any one sample.  So why is my daily graph, 
which is allegedly graphing current samples, showing 0.4?

> What you want is something other than an average.  RRD does averages.

For summaries, yes.  This isn't a summary I'm talking about.

Or is there some hidden undocumented averaging going on even for the current 
sample of the daily graph?

Still confused,

Greg.

-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com

Gmane