Temperature graphs
list Kevin Hanrahan
I cannot get the temperature graphs to work. I have added the folowing after the hostname: LARRD:*,temperature Did I make an error? What is the correct syntax please? Thanks Kevin Note: The information contained in this email and in any attachments is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. The recipient should check this email and any attachments for the presence of viruses. Sender accepts no liability for any damages caused by any virus transmitted by this email. If you have received this email in error, please notify us immediately by replying to the message and delete the email from your computer. This e-mail is and any response to it will be unencrypted and, therefore, potentially unsecure. Thank you. NOVA Information Systems, Inc.
list Charles Jones
I noted that although I did a "make setup", the new hobbitgraph.cfg file did not get copied over, which has the new temperature graph setttings in it. Try copying over that new file. -Charles
▸
user-fd47fec4b039@xymon.invalid wrote:
I cannot get the temperature graphs to work. I have added the folowing after the hostname: LARRD:*,temperature Did I make an error? What is the correct syntax please? Thanks Kevin
list Henrik Størner
▸
On Mon, Feb 14, 2005 at 04:45:11PM -0500, user-fd47fec4b039@xymon.invalid wrote:
I cannot get the temperature graphs to work. I have added the folowing after the hostname: LARRD:*,temperature
You shouldn't have to add anything - the "temperature" graph is included in the default set of graphs.
Did I make an error? What is the correct syntax please?
Your syntax is OK, the problem lies elsewhere. For starters, are your temperature reports using the name "temperature" for the status column? Henrik
list Kevin Hanrahan
Yes, the column says temperature
▸
-----Original Message-----
From: Henrik Stoerner [mailto:user-ce4a2c883f75@xymon.invalid]
Sent: Monday, February 14, 2005 4:56 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature graphs
On Mon, Feb 14, 2005 at 04:45:11PM -0500, user-fd47fec4b039@xymon.invalid wrote:I cannot get the temperature graphs to work. I have added the folowing after the hostname: LARRD:*,temperature
You shouldn't have to add anything - the "temperature" graph is included in the default set of graphs.
Did I make an error? What is the correct syntax please?
Your syntax is OK, the problem lies elsewhere. For starters, are your temperature reports using the name "temperature" for the status column? Henrik Note: The information contained in this email and in any attachments is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. The recipient should check this email and any attachments for the presence of viruses. Sender accepts no liability for any damages caused by any virus transmitted by this email. If you have received this email in error, please notify us immediately by replying to the message and delete the email from your computer. This e-mail is and any response to it will be unencrypted and, therefore, potentially unsecure. Thank you. NOVA Information Systems, Inc.
list Henrik Størner
In <user-5bd5a99fbc0f@xymon.invalid> user-fd47fec4b039@xymon.invalid writes:
Yes, the column says temperature
What does one of these reports look like ? Do you have any temperature.rrd files in the ~/data/rrd/HOSTNAME/ directory ? Regards, Henrik
list Kevin Hanrahan
Where is the new hobbitgraph.cfg file?
▸
From: Charles Jones [mailto:user-e86b4aeade4e@xymon.invalid] Sent: Monday, February 14, 2005 4:56 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature graphs
Importance: Low
▸
I noted that although I did a "make setup", the new hobbitgraph.cfg file did
not get copied over, which has the new temperature graph setttings in it.
Try copying over that new file.
-Charles
user-fd47fec4b039@xymon.invalid <mailto:user-fd47fec4b039@xymon.invalid> wrote:
I cannot get the temperature graphs to work. I have added the folowing after
the hostname:
LARRD:*,temperature
Did I make an error? What is the correct syntax please?
Thanks
Kevin
list Henrik Størner
▸
On Mon, Feb 14, 2005 at 08:41:54PM -0500, Kevin Hanrahan wrote:
Where is the new hobbitgraph.cfg file?
In the hobbit-4.0-RC2/hobbitd/etcfiles/ directory. Henrik
list Kevin Hanrahan
Henrik, There are temperature.rrd files in the ~/data/rrd/HOSTNAME/ directory but I do not get any graphs. I do get a broken "hobbit graph temperature" link however in the trends column. This is what one of the reports looks like: Device Temp(C) Temp(F) AMBIENT 17 62 CPU0 40 104 CPU1 40 104 CPU2 40 104 CPU3 40 104 Status green: All devices look okay Kevin
▸
-----Original Message-----
From: Henrik Storner [mailto:user-ce4a2c883f75@xymon.invalid]
Sent: Monday, February 14, 2005 6:25 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature graphs
In <user-5bd5a99fbc0f@xymon.invalid>
user-fd47fec4b039@xymon.invalid writes:
Yes, the column says temperature
What does one of these reports look like ? Do you have any temperature.rrd files in the ~/data/rrd/HOSTNAME/ directory ? Regards, Henrik Note: The information contained in this email and in any attachments is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. The recipient should check this email and any attachments for the presence of viruses. Sender accepts no liability for any damages caused by any virus transmitted by this email. If you have received this email in error, please notify us immediately by replying to the message and delete the email from your computer. This e-mail is and any response to it will be unencrypted and, therefore, potentially unsecure. Thank you. NOVA Information Systems, Inc.
list Henrik Størner
▸
On Tue, Feb 15, 2005 at 09:48:46AM -0500, user-fd47fec4b039@xymon.invalid wrote:
Henrik, There are temperature.rrd files in the ~/data/rrd/HOSTNAME/ directory but I do not get any graphs. I do get a broken "hobbit graph temperature" link however in the trends column.
Your status report also looks OK, so the problem is with the graph
CGI. Could you check if the following is in etc/hobbitgraph.cfg:
[temperature]
FNPATTERN temperature.(.*).rrd
TITLE Temperature
YAXIS Celsius
DEF:p at RRDIDX@=@RRDFN@:temperature:AVERAGE
LINE2:p at RRDIDX@#@COLOR@:@RRDPARAM@
GPRINT:p at RRDIDX@:LAST: \: %5.1lf (cur)
GPRINT:p at RRDIDX@:MAX: \: %5.1lf (max)
GPRINT:p at RRDIDX@:MIN: \: %5.1lf (min)
GPRINT:p at RRDIDX@:AVERAGE: \: %5.1lf (avg)\n
Henrik
list Kevin Hanrahan
That section for temperature does not exist. I saw another post from Charles Jones saying that the hobbitgraph.cfg did not get copied during a "make setup" so I tried copying that file from where you specified (hobbit-4.0-RC2/hobbitd/etcfiles/) and now the temperature graphs are there! It appears that there is a problem during install with that file (hobbitgraph.cfg). Thanks for the help guys. Kevin
▸
-----Original Message-----
From: Henrik Stoerner [mailto:user-ce4a2c883f75@xymon.invalid]
Sent: Tuesday, February 15, 2005 10:00 AM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature graphs
On Tue, Feb 15, 2005 at 09:48:46AM -0500, user-fd47fec4b039@xymon.invalid wrote:Henrik, There are temperature.rrd files in the ~/data/rrd/HOSTNAME/ directory but I do not get any graphs. I do get a broken "hobbit graph temperature" link however in the trends column.
Your status report also looks OK, so the problem is with the graph CGI.
Could you check if the following is in etc/hobbitgraph.cfg:
[temperature]
FNPATTERN temperature.(.*).rrd
TITLE Temperature
YAXIS Celsius
DEF:p at RRDIDX@=@RRDFN@:temperature:AVERAGE
LINE2:p at RRDIDX@#@COLOR@:@RRDPARAM@
GPRINT:p at RRDIDX@:LAST: \: %5.1lf (cur)
GPRINT:p at RRDIDX@:MAX: \: %5.1lf (max)
GPRINT:p at RRDIDX@:MIN: \: %5.1lf (min)
GPRINT:p at RRDIDX@:AVERAGE: \: %5.1lf (avg)\n
Henrik
Note: The information contained in this email and in any attachments is
intended only for the person or entity to which it is addressed and may
contain confidential and/or privileged material. Any review,
retransmission, dissemination or other use of, or taking of any action in
reliance upon, this information by persons or entities other than the
intended recipient is prohibited. The recipient should check this email and
any attachments for the presence of viruses. Sender accepts no liability
for any damages caused by any virus transmitted by this email. If you have
received this email in error, please notify us immediately by replying to
the message and delete the email from your computer. This e-mail is and any
response to it will be unencrypted and, therefore, potentially unsecure.
Thank you. NOVA Information Systems, Inc.
list Charles Jones
1. How can I graph Farenheit instead of Celcius? 2. I have two different temperature outputs (different machines), will Hobbit pick up the correct value from both? Here are examples, directly cut and pasted from the hobbit status page... Machine 1: (I think this is the "usual" temperature output) Tue Feb 15 11:45:33 MST 2005 Temperature status: Device Temp(C) Temp(F) AMBIENT 20 68
▸
CPU0 40 104
CPU1 40 104
CPU2 40 104
CPU3 40 104
Status green: All devices look okay
Machine 2: (This one looks a bit different)
Tue Feb 15 11:46:20 MST 2005 Temperature status:
Device Temp(C) Temp(F)
Board 0 34 93
Board 1 35 95
Board 2 35 95
Board 4 34 93
Board 6 37 98
Board CLK 25 77
Status green: All devices look okay
Note that for Machine 2 I believe the Board CLK is the value we want to graph.
list Charles Jones
Doh! Now that I actually have the graphs going, I see that Hobbit is graphing all of the values, which is even better! So now my only question is how to graph Farenheit instead of Celcius....I assume I would have to edit do_temperature.c:
sprintf(rrdvalues, "%d:%d", (int)tstamp, tmpC);
dprintf("RRD %s, value %d\n", rrdfn, tmpC);
and change those tmpC to tmpF (And also edit hobbitgraph.cfg and s/Celcius/Farenheight)?
-Charles
▸
Charles Jones wrote:
1. How can I graph Farenheit instead of Celcius? 2. I have two different temperature outputs (different machines), will Hobbit pick up the correct value from both? Here are examples, directly cut and pasted from the hobbit status page... Machine 1: (I think this is the "usual" temperature output) Tue Feb 15 11:45:33 MST 2005 Temperature status: Device Temp(C) Temp(F) AMBIENT 20 68 CPU0 40 104 CPU1 40 104 CPU2 40 104 CPU3 40 104 Status green: All devices look okay Machine 2: (This one looks a bit different) Tue Feb 15 11:46:20 MST 2005 Temperature status: Device Temp(C) Temp(F) Board 0 34 93 Board 1 35 95 Board 2 35 95 Board 4 34 93 Board 6 37 98 Board CLK 25 77 Status green: All devices look okay Note that for Machine 2 I believe the Board CLK is the value we want to graph.
list Charles Jones
I also discovered by poking around, that I can specify to hobbitgraph to only display one particular value... I noticed when you click a disk graph, that the url indicates which partitions to display via "&first=1&count=4", would show all four disk partitions. If you change that to &first=1&count=1, it only graphs the first disk partition (usually /). I tried the same thing with the temperature graphs and it worked, ie: http://www.myhobbitserver.com/hobbit-cgi/hobbitgraph.sh?host=somehost.com&service=temperature&first=1&count=1&disp=somehost%2enet&action=menu Displayed the graph for only "Ambient". I guess a question I would have about that is how could I set that as default, or even better, if there was some GUI way to select which values to include in the graph (maybe its part of Cacti already, I dunno). Sorry if this info is already in the docs somewhere, I thought it was useful info so I posted it when I discovered it. -Charles
▸
Charles Jones wrote:
Doh! Now that I actually have the graphs going, I see that Hobbit is graphing all of the values, which is even better! So now my only question is how to graph Farenheit instead of Celcius....I assume I would have to edit do_temperature.c: sprintf(rrdvalues, "%d:%d", (int)tstamp, tmpC); dprintf("RRD %s, value %d\n", rrdfn, tmpC); and change those tmpC to tmpF (And also edit hobbitgraph.cfg and s/Celcius/Farenheight)? -Charles Charles Jones wrote:1. How can I graph Farenheit instead of Celcius? 2. I have two different temperature outputs (different machines), will Hobbit pick up the correct value from both? Here are examples, directly cut and pasted from the hobbit status page... Machine 1: (I think this is the "usual" temperature output) Tue Feb 15 11:45:33 MST 2005 Temperature status: Device Temp(C) Temp(F) AMBIENT 20 68 CPU0 40 104 CPU1 40 104 CPU2 40 104 CPU3 40 104 Status green: All devices look okay Machine 2: (This one looks a bit different) Tue Feb 15 11:46:20 MST 2005 Temperature status: Device Temp(C) Temp(F) Board 0 34 93 Board 1 35 95 Board 2 35 95 Board 4 34 93 Board 6 37 98 Board CLK 25 77 Status green: All devices look okay Note that for Machine 2 I believe the Board CLK is the value we want to graph.
list Henrik Størner
▸
On Tue, Feb 15, 2005 at 11:49:55AM -0700, Charles Jones wrote:
1. How can I graph Farenheit instead of Celcius?
The easiest way of doing that is to change the hobbitgraph.cfg
definition for the graph!
In the [temperature] section, there's this line that picks out the
"temperature" setting from the RRD file:
DEF:p at RRDIDX@=@RRDFN@:temperature:AVERAGE
This is an RRDtool "DEF" (definition) that defines a variable
"p at RRDIDX@" (the "@RRDIDX@" is handled by hobbitgraph, and is needed
because there are multiple graphs. So essentially you get p0, p1, p2
etc - one for each temperature*rrd file).
There's also something called a "CDEF" - a Computed DEFinition.
Meaning you can do math on the data you have in the RRD. So to compute
Fahrenheit, you do the usual F = (9/5)*C + 32 - using Reverse Polish
Notation (RPN) which is what CDEF's understand:
CDEF:tempf at RRDIDX@=9,5,/,p at RRDIDX@,*,32,+
Now you have two values: p at RRDIDX which is the original Celsius
temperature, and tempf at RRDIDX@ which is the Fahrenheit equivalent.
So to get the graph for the Fahrenheit one, change the LINE2 graph
definition to use the Fahrenheit value:
LINE2:tempf at RRDIDX@#@COLOR@:@RRDPARAM@
Finally you just change the Y-axis label, and end up with this
definition:
[temperature]
FNPATTERN temperature.(.*).rrd
TITLE Temperature
YAXIS Fahrenheit
DEF:p at RRDIDX@=@RRDFN@:temperature:AVERAGE
CDEF:tempf at RRDIDX@=9,5,/,p at RRDIDX@,*,32,+
LINE2:tempf at RRDIDX@#@COLOR@:@RRDPARAM@
▸
GPRINT:p at RRDIDX@:LAST: \: %5.1lf (cur)
GPRINT:p at RRDIDX@:MAX: \: %5.1lf (max)
GPRINT:p at RRDIDX@:MIN: \: %5.1lf (min)
GPRINT:p at RRDIDX@:AVERAGE: \: %5.1lf (avg)\n
Untested, but you get the idea ... the RRDtool website has some more
examples and tutorials on CDEF's.
Regards,
Henrik
list Charles Jones
Thanks Henrik, as usual your advice worked like a charm. I also changed the GPRINT statements to use tempf so the legend would be correct. Final version:
▸
[temperature]
FNPATTERN temperature.(.*).rrd
TITLE Temperature
YAXIS Fahrenheit
DEF:p at RRDIDX@=@RRDFN@:temperature:AVERAGE
CDEF:tempf at RRDIDX@=9,5,/,p at RRDIDX@,*,32,+
LINE2:tempf at RRDIDX@#@COLOR@:@RRDPARAM@
GPRINT:tempf at RRDIDX@:LAST: \: %5.1lf (cur)
GPRINT:tempf at RRDIDX@:MAX: \: %5.1lf (max)
GPRINT:tempf at RRDIDX@:MIN: \: %5.1lf (min)
GPRINT:tempf at RRDIDX@:AVERAGE: \: %5.1lf (avg)\n
Thanks again!
-Charles
▸
Henrik Stoerner wrote:
On Tue, Feb 15, 2005 at 11:49:55AM -0700, Charles Jones wrote:1. How can I graph Farenheit instead of Celcius?The easiest way of doing that is to change the hobbitgraph.cfg definition for the graph! In the [temperature] section, there's this line that picks out the "temperature" setting from the RRD file: DEF:p at RRDIDX@=@RRDFN@:temperature:AVERAGE This is an RRDtool "DEF" (definition) that defines a variable "p at RRDIDX@" (the "@RRDIDX@" is handled by hobbitgraph, and is needed because there are multiple graphs. So essentially you get p0, p1, p2 etc - one for each temperature*rrd file). There's also something called a "CDEF" - a Computed DEFinition. Meaning you can do math on the data you have in the RRD. So to compute Fahrenheit, you do the usual F = (9/5)*C + 32 - using Reverse Polish Notation (RPN) which is what CDEF's understand: CDEF:tempf at RRDIDX@=9,5,/,p at RRDIDX@,*,32,+ Now you have two values: p at RRDIDX which is the original Celsius temperature, and tempf at RRDIDX@ which is the Fahrenheit equivalent. So to get the graph for the Fahrenheit one, change the LINE2 graph definition to use the Fahrenheit value: LINE2:tempf at RRDIDX@#@COLOR@:@RRDPARAM@ Finally you just change the Y-axis label, and end up with this definition: [temperature] FNPATTERN temperature.(.*).rrd TITLE Temperature YAXIS Fahrenheit DEF:p at RRDIDX@=@RRDFN@:temperature:AVERAGE CDEF:tempf at RRDIDX@=9,5,/,p at RRDIDX@,*,32,+ LINE2:tempf at RRDIDX@#@COLOR@:@RRDPARAM@ GPRINT:p at RRDIDX@:LAST: \: %5.1lf (cur) GPRINT:p at RRDIDX@:MAX: \: %5.1lf (max) GPRINT:p at RRDIDX@:MIN: \: %5.1lf (min) GPRINT:p at RRDIDX@:AVERAGE: \: %5.1lf (avg)\n Untested, but you get the idea ... the RRDtool website has some more examples and tutorials on CDEF's. Regards, Henrik
list Henrik Størner
▸
On Tue, Feb 15, 2005 at 01:04:54PM -0700, Charles Jones wrote:
Thanks Henrik, as usual your advice worked like a charm. I also changed
the GPRINT statements to use tempf so the legend would be correct.
Final version:
[temperature]
FNPATTERN temperature.(.*).rrd
TITLE Temperature
YAXIS Fahrenheitthanks, I've included this in the default hobbitgraph.cfg file - commented out, but it will be easy to enable. Regards, Henrik
list Henrik Størner
▸
On Tue, Feb 15, 2005 at 12:07:00PM -0700, Charles Jones wrote:
I noticed when you click a disk graph, that the url indicates which partitions to display via "&first=1&count=4", would show all four disk partitions. If you change that to &first=1&count=1, it only graphs the first disk partition (usually /). I tried the same thing with the temperature graphs and it worked, ie: http://www.myhobbitserver.com/hobbit-cgi/hobbitgraph.sh?host=somehost.com&service=temperature&first=1&count=1&disp=somehost%2enet&action=menu Displayed the graph for only "Ambient".
You have sharp eyes - yes, this is precisely how hobbitgraph handles cases where there are multiple graphs combined in one image. It builds a list of the available graphs, sort them by a key - for disks that is the filesystem name, for temperature it's the name of the temperature monitor - and uses the "first" and "count" settings to decide which ones from this set of data to include on the current graph.
▸
I guess a question I would have about that is how could I set that as default, or even better, if there was some GUI way to select which values to include in the graph (maybe its part of Cacti already, I dunno).
There's currently no way of setting up a default view of a graph - apart from which graph to include in the default set, you might also want the monthly graph as the default, instead of the 48-hour graph; or you want the graph to be clipped at some maximum value by default; or ... lots of other possibilities. Plus a GUI to dynamically tune the graph in real-time. It's something I want to improve after the 4.0 release is out. So I'll happily note your input, but postpone any implementation of it till later. Some of the stuff I want to do requires changes to the RRDtool library - Tobi Oetiker (RRDtool author) doesn't want to do these changes for RRDtool 1.0.x, so I'll probably end up including a modified RRDtool with Hobbit. And that's just too much work for 4.0. Regards, Henrik
list Patrick Nixon
Hey all, I'm trying to monitor some temperatures from Windows boxes and I have the external script set up to send the message to Xymon, and it's trying to create the graph, but It's not seeing any data, even over the course of several hours worth of data. Does anyone know the exact formatting necessary to create the graphs? Also, is it possible to graph multiple items on one graph? Thanks! --Pat
list dOCtoR MADneSs
▸
Patrick Nixon a écrit :
Hey all, I'm trying to monitor some temperatures from Windows boxes and I have the external script set up to send the message to Xymon, and it's trying to create the graph, but It's not seeing any data, even over the course of several hours worth of data. Does anyone know the exact formatting necessary to create the graphs? Also, is it possible to graph multiple items on one graph? Thanks! --Pat
Hi,
For your question about multiple items on one graph, you can do the
followin instructions (from hobbitgraphs.cfg)
# For graphs that use multiple RRD files as input, you specify a filename
# pattern in the FNPATTERN statement, and optionally a pattern of files
# to exclude from the graph with EXFNPATTERN (see "[tcp]" for an example).
# When FNPATTERN is used, you can use "@RRDFN@" in the RRDtool definitions
# to pick up each filename. "@RRDIDX@" is an index (starting at 0) for each
# file in the set. "@RRDPARAM@" contains the first word extracted from the
# pattern of files (see e.g. "[memory]" how this is used). "@COLOR@" picks
# a new color for each graph automatically.
So, all you need is to have items in separates RRD files, using
SPLITNCV_test instead of NCV_test in hobbitserver.cfg file.
Then follow the instructions.
My running configuration for this kind of graph (multiple items in 1
graph) is :
[hardware]
TITLE Temperatures
YAXIS Celsius
FNPATTERN hardware,(.*)_temperature.rrd
DEF:@RRDIDX@=@RRDFN@:lambda:AVERAGE
LINE2:@RRDIDX@#@COLOR@:@RRDPARAM@
GPRINT:@RRDIDX@:LAST: \: %5.1lf (cur)
GPRINT:@RRDIDX@:MAX: \: %5.1lf (max)
GPRINT:@RRDIDX@:MIN: \: %5.1lf (min)
GPRINT:@RRDIDX@:AVERAGE: \: %5.1lf (avg)\n
list Stef Coene
▸
On Sunday 25 January 2009, Patrick Nixon wrote:
Hey all, I'm trying to monitor some temperatures from Windows boxes and I have the external script set up to send the message to Xymon, and it's trying to create the graph, but It's not seeing any data, even over the course of several hours worth of data. Does anyone know the exact formatting necessary to create the graphs?
It depends. Are you using NCV or are you sending 'trends' messages?
Also, is it possible to graph multiple items on one graph?
Yes. Stef
list Patrick Nixon
See, this is where I admit I didn't know you had to do that. I thought this was one of those services (like disk/cpu) where it pulled it out of the data automatically. I'll look into the two types (NCV/split ncv and trends) --Pat
▸
On Sun, Jan 25, 2009 at 1:38 PM, Stef Coene <user-dbffe946c0f4@xymon.invalid> wrote:On Sunday 25 January 2009, Patrick Nixon wrote:Hey all, I'm trying to monitor some temperatures from Windows boxes and I have the external script set up to send the message to Xymon, and it's trying to create the graph, but It's not seeing any data, even over the course of several hours worth of data. Does anyone know the exact formatting necessary to create the graphs?It depends. Are you using NCV or are you sending 'trends' messages?Also, is it possible to graph multiple items on one graph?Yes. Stef
list Greg L Hubbard
Make sure you are logging the data as a guage, and that RRD is not trying to derive readings by comparing the current reading with the prior reading... GLH
▸
-----Original Message-----
From: Stef Coene [mailto:user-dbffe946c0f4@xymon.invalid] Sent: Sunday, January 25, 2009 12:38 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature Graphs
On Sunday 25 January 2009, Patrick Nixon wrote:Hey all, I'm trying to monitor some temperatures from Windows boxes and I have
the external script set up to send the message to Xymon, and it's trying to create the graph, but It's not seeing any data, even over the course of several hours worth of data. Does anyone know the exact formatting necessary to create the graphs?
It depends. Are you using NCV or are you sending 'trends' messages?
Also, is it possible to graph multiple items on one graph?
Yes. Stef
list Patrick Nixon
okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!
list Gavin Leonard
How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin
▸
-----Original Message-----
From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid]
Sent: Monday, January 26, 2009 7:54 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature Graphs
okay,
I figured out that Xymon can handle the temperature graph
automatically given the correct data format.
I found documented format to be:
Device Temp(C) Temp(F)
&green Temp1 42 90
&green Temp2 34 80
&green ACPI-Temp1 45 92
It's working all fine.
Thanks for all the help!
list Ralph Mitchell
If the systems don't already have their own internal heat sensors, you can easily add a One-Wire sensor to the closet. You can get ready-built modules here: http://www.hobby-boards.com/catalog/product_info.php?cPath=24&products_id=109&osCsid=efa67c007302cef289ca6efa1559e603 and either the serial or the USB adaptor: http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=28 http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=1503 Software to read One-Wire devices is here: http://www.digitemp.com/software.shtml You'll also need their cable to connect the module to the adaptor. If you're handy with a soldering iron you could skip the module and just get the bare temperature sensor (DS18S20), cut one jack off a phone cord and solder the sensor to the relevant wires. The other jack on the phone wire will plug directly into the adaptor. I've got a script for reading the temps via digitemp and reporting to Xymon, but the machine it's on is currently unreachable. I'm not affiliated in any way to Hobby Boards, not even as a customer, but I do have some of those sensors monitoring temps in parts of my workshop. Ralph Mitchell
▸
On Tue, Jan 27, 2009 at 2:34 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid>wrote:
How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!
list Gavin Leonard
These are all new HP DL series servers so they should have that capability already.. just need to configure.. -Gavin
▸
From: Ralph Mitchell [mailto:user-00a5e44c48c0@xymon.invalid] Sent: Tuesday, January 27, 2009 3:00 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs If the systems don't already have their own internal heat sensors, you can easily add a One-Wire sensor to the closet. You can get ready-built modules here: http://www.hobby-boards.com/catalog/product_info.php?cPath=24&products_id=109&osCsid=efa67c007302cef289ca6efa1559e603 and either the serial or the USB adaptor: http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=28 http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=1503 Software to read One-Wire devices is here: http://www.digitemp.com/software.shtml You'll also need their cable to connect the module to the adaptor. If you're handy with a soldering iron you could skip the module and just get the bare temperature sensor (DS18S20), cut one jack off a phone cord and solder the sensor to the relevant wires. The other jack on the phone wire will plug directly into the adaptor. I've got a script for reading the temps via digitemp and reporting to Xymon, but the machine it's on is currently unreachable. I'm not affiliated in any way to Hobby Boards, not even as a customer, but I do have some of those sensors monitoring temps in parts of my workshop. Ralph Mitchell On Tue, Jan 27, 2009 at 2:34 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid<mailto:user-d65663809eb4@xymon.invalid>> wrote: How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid<mailto:user-1f2406f832af@xymon.invalid>] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid<mailto:user-ae9b8668bcde@xymon.invalid> Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!
list Ralph Mitchell
Compaq/HP used to have a Linux package for DL380 server health, but that was for the 2.4 kernel and Redhat 7.2 about 10 years ago. What's the OS?? Someone here might have prior experience.
▸
Ralph Mitchell
On Tue, Jan 27, 2009 at 5:45 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid>wrote:
These are all new HP DL series servers so they should have that capability already.. just need to configure.. -Gavin *From:* Ralph Mitchell [mailto:user-00a5e44c48c0@xymon.invalid] *Sent:* Tuesday, January 27, 2009 3:00 PM *To:* user-ae9b8668bcde@xymon.invalid *Subject:* Re: [hobbit] Temperature Graphs If the systems don't already have their own internal heat sensors, you can easily add a One-Wire sensor to the closet. You can get ready-built modules here: http://www.hobby-boards.com/catalog/product_info.php?cPath=24&products_id=109&osCsid=efa67c007302cef289ca6efa1559e603 and either the serial or the USB adaptor: http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=28 http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=1503 Software to read One-Wire devices is here: http://www.digitemp.com/software.shtml You'll also need their cable to connect the module to the adaptor. If you're handy with a soldering iron you could skip the module and just get the bare temperature sensor (DS18S20), cut one jack off a phone cord and solder the sensor to the relevant wires. The other jack on the phone wire will plug directly into the adaptor. I've got a script for reading the temps via digitemp and reporting to Xymon, but the machine it's on is currently unreachable. I'm not affiliated in any way to Hobby Boards, not even as a customer, but I do have some of those sensors monitoring temps in parts of my workshop. Ralph Mitchell
On Tue, Jan 27, 2009 at 2:34 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid>
▸
wrote:
How is this done? I have some systems in a closet that we have problems
with heat and I would love to be able to monitor this with hobbit..
-Gavin
-----Original Message-----
From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid]
Sent: Monday, January 26, 2009 7:54 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature Graphs
okay,
I figured out that Xymon can handle the temperature graph
automatically given the correct data format.
I found documented format to be:
Device Temp(C) Temp(F)
&green Temp1 42 90
&green Temp2 34 80
&green ACPI-Temp1 45 92
It's working all fine.
Thanks for all the help!
list Patrick Nixon
The boxes I'm monitoring are windows box, so I had to go a slightly convulted route to get the temperature information 1.) Install SpeedFan to enable monitoring for the desired temperature probes 2.) Configure SpeedFan to run as a service at boot up to enable non-login monitoring 3.) Installed ActiveState Perl to run the script to extract the temperature information out of SpeedFan's log file and output the file into BBNT's saved logs directory 4.) Configured Windows to run the temp perl script every five minutes In terms of Linux servers, I know there are scripts out there that will work on lm-sensors and you may have to tweak them to get the right output, but I don't imagine it will be hard. I just don't have any linux boxes in this setup that have a working lm-sensors setup. If you'd like the script I wrote, even though it's probably horrible perl, I can post it up. --Pat
▸
On Tue, Jan 27, 2009 at 3:34 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid> wrote:How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!
list dOCtoR MADneSs
▸
On Wed, 28 Jan 2009 10:33:42 -0500, Patrick Nixon <user-1f2406f832af@xymon.invalid> wrote:
The boxes I'm monitoring are windows box, so I had to go a slightly convulted route to get the temperature information 1.) Install SpeedFan to enable monitoring for the desired temperature probes 2.) Configure SpeedFan to run as a service at boot up to enable non-login monitoring 3.) Installed ActiveState Perl to run the script to extract the temperature information out of SpeedFan's log file and output the file into BBNT's saved logs directory 4.) Configured Windows to run the temp perl script every five minutes In terms of Linux servers, I know there are scripts out there that will work on lm-sensors and you may have to tweak them to get the right output, but I don't imagine it will be hard. I just don't have any linux boxes in this setup that have a working lm-sensors setup. If you'd like the script I wrote, even though it's probably horrible perl, I can post it up. --Pat On Tue, Jan 27, 2009 at 3:34 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid> wrote:How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!
On my side, I'm monitoring almost only linuxes. I wrote a script to get temp/speed/voltages from sensors using lm-sensors (and hddtemp for hdd temp...) it's tested on a few different machines (like 4), but seems to be not too much buggy. You can get it here : https://www.makelofine.org/xymon-plugins/hobbit-hardware-v0.1.tar.bz2 If you take it, just tell me if you changed something in, so i'll include your changes.
list Michael Nemeth
Your scripts are NOT posting correctly on the Shire. Yes you've a link, which for me is blocked at work and at home was not working either (forget why).
▸
user-d54077869176@xymon.invalid wrote:On Wed, 28 Jan 2009 10:33:42 -0500, Patrick Nixon <user-1f2406f832af@xymon.invalid> wrote:The boxes I'm monitoring are windows box, so I had to go a slightly convulted route to get the temperature information 1.) Install SpeedFan to enable monitoring for the desired temperature probes 2.) Configure SpeedFan to run as a service at boot up to enable non-login monitoring 3.) Installed ActiveState Perl to run the script to extract the temperature information out of SpeedFan's log file and output the file into BBNT's saved logs directory 4.) Configured Windows to run the temp perl script every five minutes In terms of Linux servers, I know there are scripts out there that will work on lm-sensors and you may have to tweak them to get the right output, but I don't imagine it will be hard. I just don't have any linux boxes in this setup that have a working lm-sensors setup. If you'd like the script I wrote, even though it's probably horrible perl, I can post it up. --Pat On Tue, Jan 27, 2009 at 3:34 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid> wrote:How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!On my side, I'm monitoring almost only linuxes. I wrote a script to get temp/speed/voltages from sensors using lm-sensors (and hddtemp for hdd temp...) it's tested on a few different machines (like 4), but seems to be not too much buggy. You can get it here : https://www.makelofine.org/xymon-plugins/hobbit-hardware-v0.1.tar.bz2 If you take it, just tell me if you changed something in, so i'll include your changes.
list dOCtoR MADneSs
Can you tell me more about what's wrong with them ? If you mean about output format, I created my own, and create graph definition for it. If you mean anything else, please tell me more, I'll see to correct them. On Fri, 30 Jan 2009 08:08:44 -0500, michael nemeth
▸
<user-609d3fab5b2d@xymon.invalid> wrote:Your scripts are NOT posting correctly on the Shire. Yes you've a link, which for me is blocked at work and at home was not working either (forget why). user-d54077869176@xymon.invalid wrote:On Wed, 28 Jan 2009 10:33:42 -0500, Patrick Nixon <user-1f2406f832af@xymon.invalid> wrote:The boxes I'm monitoring are windows box, so I had to go a slightly convulted route to get the temperature information 1.) Install SpeedFan to enable monitoring for the desired temperature probes 2.) Configure SpeedFan to run as a service at boot up to enable non-login monitoring 3.) Installed ActiveState Perl to run the script to extract the temperature information out of SpeedFan's log file and output the file into BBNT's saved logs directory 4.) Configured Windows to run the temp perl script every five minutes In terms of Linux servers, I know there are scripts out there that will work on lm-sensors and you may have to tweak them to get the right output, but I don't imagine it will be hard. I just don't have any linux boxes in this setup that have a working lm-sensors setup. If you'd like the script I wrote, even though it's probably horrible perl, I can post it up. --Pat On Tue, Jan 27, 2009 at 3:34 PM, Gavin Leonard
<user-d65663809eb4@xymon.invalid>
wrote:How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!On my side, I'm monitoring almost only linuxes. I wrote a script to get temp/speed/voltages from sensors using lm-sensors (and hddtemp for hdd temp...) it's tested on a few different machines (like 4), but seems to be not too much buggy. You can get it here : https://www.makelofine.org/xymon-plugins/hobbit-hardware-v0.1.tar.bz2 If you take it, just tell me if you changed something in, so i'll include your changes.
list Michael Nemeth
the SCRIPTS maybe fine; but they are not format on the shire un readable.
▸
user-d54077869176@xymon.invalid wrote:Can you tell me more about what's wrong with them ? If you mean about output format, I created my own, and create graph definition for it. If you mean anything else, please tell me more, I'll see to correct them. On Fri, 30 Jan 2009 08:08:44 -0500, michael nemeth <user-609d3fab5b2d@xymon.invalid> wrote:Your scripts are NOT posting correctly on the Shire. Yes you've a link, which for me is blocked at work and at home was not working either (forget why). user-d54077869176@xymon.invalid wrote:On Wed, 28 Jan 2009 10:33:42 -0500, Patrick Nixon <user-1f2406f832af@xymon.invalid> wrote:The boxes I'm monitoring are windows box, so I had to go a slightly convulted route to get the temperature information 1.) Install SpeedFan to enable monitoring for the desired temperature probes 2.) Configure SpeedFan to run as a service at boot up to enable non-login monitoring 3.) Installed ActiveState Perl to run the script to extract the temperature information out of SpeedFan's log file and output the file into BBNT's saved logs directory 4.) Configured Windows to run the temp perl script every five minutes In terms of Linux servers, I know there are scripts out there that will work on lm-sensors and you may have to tweak them to get the right output, but I don't imagine it will be hard. I just don't have any linux boxes in this setup that have a working lm-sensors setup. If you'd like the script I wrote, even though it's probably horrible perl, I can post it up. --Pat On Tue, Jan 27, 2009 at 3:34 PM, Gavin Leonard<user-d65663809eb4@xymon.invalid>wrote:How is this done? I have some systems in a closet that we haveproblemswith heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!On my side, I'm monitoring almost only linuxes. I wrote a script to get temp/speed/voltages from sensors using lm-sensors (and hddtemp for hdd temp...) it's tested on a few different machines (like 4), but seems to be nottoomuch buggy. You can get it here : https://www.makelofine.org/xymon-plugins/hobbit-hardware-v0.1.tar.bz2 If you take it, just tell me if you changed something in, so i'llincludeyour changes.
list dOCtoR MADneSs
Tks, I didn't see it, I corrected using <code sh>, is it ok now ? On Fri, 30 Jan 2009 11:47:35 -0500, michael nemeth
▸
<user-609d3fab5b2d@xymon.invalid> wrote:the SCRIPTS maybe fine; but they are not format on the shire un readable. user-d54077869176@xymon.invalid wrote:Can you tell me more about what's wrong with them ? If you mean about output format, I created my own, and create graph definition for it. If you mean anything else, please tell me more, I'll see to correct them. On Fri, 30 Jan 2009 08:08:44 -0500, michael nemeth <user-609d3fab5b2d@xymon.invalid> wrote:Your scripts are NOT posting correctly on the Shire. Yes you've a link,
which for me is blocked at work and at home was not working either (forget why). user-d54077869176@xymon.invalid wrote:On Wed, 28 Jan 2009 10:33:42 -0500, Patrick Nixon <user-1f2406f832af@xymon.invalid> wrote:The boxes I'm monitoring are windows box, so I had to go a slightly convulted route to get the temperature information 1.) Install SpeedFan to enable monitoring for the desired temperature probes 2.) Configure SpeedFan to run as a service at boot up to enable non-login monitoring 3.) Installed ActiveState Perl to run the script to extract the temperature information out of SpeedFan's log file and output the file into BBNT's saved logs directory 4.) Configured Windows to run the temp perl script every five minutes In terms of Linux servers, I know there are scripts out there that will work on lm-sensors and you may have to tweak them to get the right output, but I don't imagine it will be hard. I just don't have any linux boxes in this setup that have a working lm-sensors setup. If you'd like the script I wrote, even though it's probably horrible perl, I can post it up. --Pat On Tue, Jan 27, 2009 at 3:34 PM, Gavin Leonard<user-d65663809eb4@xymon.invalid>wrote:How is this done? I have some systems in a closet that we haveproblemswith heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!On my side, I'm monitoring almost only linuxes. I wrote a script to get temp/speed/voltages from sensors using lm-sensors (and hddtemp for hdd temp...) it's tested on a few different machines (like 4), but seems to be nottoomuch buggy. You can get it here : https://www.makelofine.org/xymon-plugins/hobbit-hardware-v0.1.tar.bz2 If you take it, just tell me if you changed something in, so i'llincludeyour changes.
list Gavin Leonard
They are a mixture of windows 03/08 and centos 5.x.. all servers are less than 3 years old.
▸
-Gavin From: Ralph Mitchell [mailto:user-00a5e44c48c0@xymon.invalid] Sent: Tuesday, January 27, 2009 9:59 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs Compaq/HP used to have a Linux package for DL380 server health, but that was for the 2.4 kernel and Redhat 7.2 about 10 years ago. What's the OS?? Someone here might have prior experience. Ralph Mitchell On Tue, Jan 27, 2009 at 5:45 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid<mailto:user-d65663809eb4@xymon.invalid>> wrote: These are all new HP DL series servers so they should have that capability already.. just need to configure.. -Gavin From: Ralph Mitchell [mailto:user-00a5e44c48c0@xymon.invalid<mailto:user-00a5e44c48c0@xymon.invalid>] Sent: Tuesday, January 27, 2009 3:00 PM To: user-ae9b8668bcde@xymon.invalid<mailto:user-ae9b8668bcde@xymon.invalid> Subject: Re: [hobbit] Temperature Graphs If the systems don't already have their own internal heat sensors, you can easily add a One-Wire sensor to the closet. You can get ready-built modules here: http://www.hobby-boards.com/catalog/product_info.php?cPath=24&products_id=109&osCsid=efa67c007302cef289ca6efa1559e603 and either the serial or the USB adaptor: http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=28 http://www.hobby-boards.com/catalog/product_info.php?cPath=23&products_id=1503 Software to read One-Wire devices is here: http://www.digitemp.com/software.shtml You'll also need their cable to connect the module to the adaptor. If you're handy with a soldering iron you could skip the module and just get the bare temperature sensor (DS18S20), cut one jack off a phone cord and solder the sensor to the relevant wires. The other jack on the phone wire will plug directly into the adaptor. I've got a script for reading the temps via digitemp and reporting to Xymon, but the machine it's on is currently unreachable. I'm not affiliated in any way to Hobby Boards, not even as a customer, but I do have some of those sensors monitoring temps in parts of my workshop. Ralph Mitchell On Tue, Jan 27, 2009 at 2:34 PM, Gavin Leonard <user-d65663809eb4@xymon.invalid<mailto:user-d65663809eb4@xymon.invalid>> wrote: How is this done? I have some systems in a closet that we have problems with heat and I would love to be able to monitor this with hobbit.. -Gavin -----Original Message----- From: Patrick Nixon [mailto:user-1f2406f832af@xymon.invalid<mailto:user-1f2406f832af@xymon.invalid>] Sent: Monday, January 26, 2009 7:54 PM To: user-ae9b8668bcde@xymon.invalid<mailto:user-ae9b8668bcde@xymon.invalid> Subject: Re: [hobbit] Temperature Graphs okay, I figured out that Xymon can handle the temperature graph automatically given the correct data format. I found documented format to be: Device Temp(C) Temp(F) &green Temp1 42 90 &green Temp2 34 80 &green ACPI-Temp1 45 92 It's working all fine. Thanks for all the help!
list Buchan Milne
▸
----- "Gavin Leonard" <user-d65663809eb4@xymon.invalid> wrote:
They are a mixture of windows 03/08 and centos 5.x.. all servers are less than 3 years old.
devmon has a compaq-server (and, compaq-servernohspare, for servers which don't have hot spare drives in the RAID config) template, which supports the HP Insight agents. The agents are available for Windows and Linux. I am monitoring a number of DL-360G4s, DL-380G4s, DL-580-G4s, DL-585s, running Windows and Linux (Red Hat Enterprise Linux 3/4/5).
In the Linux case, setup of the HP Insight agent is simply a matter of:
-installing the hpasm package
-running 'hpasm setup' (I think)
The devmon template supports:
1)RAID controller monitoring ('raid' test)
2)Fans ('fans' test)
3)Power supplies ('power' test)
4)The IML log ('log' test)
5)Temperatures ('temp' test).
The devmon rrd collector for Hobbit (now included in Xymon), supports collecting RRDs from the temp test, just include 'temp=devmon' in your TEST2RRD line in hobbitserver.cfg, and ensure that the devmon graph definitions (in extras/devmon-graph.cfg) are included in your Hobbit graph configuration, and you should get temperature graphs.
The dell-poweredge template (for Dell servers running OMSA) supports almost all the same features as well (except for IML).
BTW, I am also monitoring some Linux boxes that have lmsensor's supported devices with devmon ... but I might not have committed the test for the net-snmp lmsensors MIB yet
Regards,
Buchan
▸
From: Ralph Mitchell [mailto:user-00a5e44c48c0@xymon.invalid] Sent: Tuesday, January 27, 2009 9:59 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs Compaq/HP used to have a Linux package for DL380 server health, but that was for the 2.4 kernel and Redhat 7.2 about 10 years ago. What's the OS?? Someone here might have prior experience.
list Gavin Leonard
Awesome... thanks! -Gavin
▸
-----Original Message-----
From: Buchan Milne [mailto:user-9b139aff4dec@xymon.invalid]
Sent: Thursday, March 12, 2009 1:14 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Temperature Graphs
----- "Gavin Leonard" <user-d65663809eb4@xymon.invalid> wrote:
They are a mixture of windows 03/08 and centos 5.x.. all servers are less than 3 years old.
devmon has a compaq-server (and, compaq-servernohspare, for servers which don't have hot spare drives in the RAID config) template, which supports the HP Insight agents. The agents are available for Windows and Linux. I am monitoring a number of DL-360G4s, DL-380G4s, DL-580-G4s, DL-585s, running Windows and Linux (Red Hat Enterprise Linux 3/4/5).
In the Linux case, setup of the HP Insight agent is simply a matter of:
-installing the hpasm package
-running 'hpasm setup' (I think)
The devmon template supports:
1)RAID controller monitoring ('raid' test)
2)Fans ('fans' test)
3)Power supplies ('power' test)
4)The IML log ('log' test)
5)Temperatures ('temp' test).
The devmon rrd collector for Hobbit (now included in Xymon), supports collecting RRDs from the temp test, just include 'temp=devmon' in your TEST2RRD line in hobbitserver.cfg, and ensure that the devmon graph definitions (in extras/devmon-graph.cfg) are included in your Hobbit graph configuration, and you should get temperature graphs.
The dell-poweredge template (for Dell servers running OMSA) supports almost all the same features as well (except for IML).
BTW, I am also monitoring some Linux boxes that have lmsensor's supported devices with devmon ... but I might not have committed the test for the net-snmp lmsensors MIB yet
Regards,
Buchan
From: Ralph Mitchell [mailto:user-00a5e44c48c0@xymon.invalid] Sent: Tuesday, January 27, 2009 9:59 PM To: user-ae9b8668bcde@xymon.invalid Subject: Re: [hobbit] Temperature Graphs Compaq/HP used to have a Linux package for DL380 server health, but that was for the 2.4 kernel and Redhat 7.2 about 10 years ago. What's the OS?? Someone here might have prior experience.
list David Fogarty
I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd When I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done. Reformatting output on that may be a bit harder. The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com
list Paul Root
You'd probably have better luck if you made it colon separated. Routing Engine Temperature: 43
▸
-----Original Message-----
From: Xymon [mailto:xymon-bounces at xymon.com] On Behalf Of David Fogarty
Sent: Thursday, June 11, 2015 1:47 PM
To: 'xymon at xymon.com'
Subject: [Xymon] Temperature graphs
I've got a couple snmp tests and I'm trying to graph their output. I
tried putting in my own rrd script to collect the data, but there seems to
be a built-in one overriding mine. So, I tried to format my test output
to work with that, but I'm just getting rrd files that lack any data.
The main line of the test is like this:
&green Routing Engine Temperature 43
It causes the creation of a file like this:
/var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd
When I use "rrdtool dump" I still get data like this and an empty graph.
<!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row>
<!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row>
If someone knows the correct format for temperature tests, I'd be set.
Also, we've got snmp collecting memory data for some equipment that we'd
like to override internal graphing on specific hosts if that can be done.
Reformatting output on that may be a bit harder.
The obvious answer is to change the test names for our custom ones, but my
boss is very opposed to creating new columns in xymon.
Thanks,
Dave Fogarty
--
Senior System Administrator
CollegeNET.com
This communication is the property of CenturyLink and may contain confidential or privileged information. Unauthorized use of this communication is strictly prohibited and may be unlawful. If you have received this communication in error, please immediately notify the sender by reply e-mail and destroy all copies of the communication and any attachments.
list Bill Rowan
Get rid of the spaces. Your status should read "&green Routing_Engine_Temperature 43". Make sure you send the data in Celsius and let xymon do the conversion in the graphs.cfg file. I haven't tried using a colon for separation of data and value. I have been using just a space and it works fine for me on 4.3.7.
▸
On 6/11/2015 1:46 PM, David Fogarty wrote:I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd When I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done. Reformatting output on that may be a bit harder. The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com
list Jeremy Ruffer
I use a colon to separate value and data. It works. Jeremy
▸
On 11 Jun 2015, at 21:03, Bill Rowan <user-3078c15a8e5e@xymon.invalid> wrote: Get rid of the spaces. Your status should read "&green Routing_Engine_Temperature 43". Make sure you send the data in Celsius and let xymon do the conversion in the graphs.cfg file. I haven't tried using a colon for separation of data and value. I have been using just a space and it works fine for me on 4.3.7.On 6/11/2015 1:46 PM, David Fogarty wrote: I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd When I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done. Reformatting output on that may be a bit harder. The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com
list Galen Johnson
I'm pretty sure it's either...I find it a bit frustrating that we can't limit where it looks for that info...if you happen to have a non-related comment with a space or colon and a number, Xymon will attempt to parse it as a tracked variable as well. =G=
▸
From: Xymon <xymon-bounces at xymon.com> on behalf of Jeremy Ruffer <user-6d8e227afca3@xymon.invalid>
Sent: Thursday, June 11, 2015 4:20 PM
To: Bill Rowan
Cc: xymon at xymon.com
Subject: Re: [Xymon] Temperature graphs
I use a colon to separate value and data. It works.
Jeremy
On 11 Jun 2015, at 21:03, Bill Rowan <user-3078c15a8e5e@xymon.invalid> wrote: Get rid of the spaces. Your status should read "&green Routing_Engine_Temperature 43". Make sure you send the data in Celsius and let xymon do the conversion in the graphs.cfg file. I haven't tried using a colon for separation of data and value. I have been using just a space and it works fine for me on 4.3.7.On 6/11/2015 1:46 PM, David Fogarty wrote: I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd When I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done. Reformatting output on that may be a bit harder. The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com
list Vernon Everett
Hi all The simple solution is to send two messages. One, as a status, which is the display we get to see. And another with data, which we don't get to see, but gets parsed into the rrd files. Regards Vernon
▸
On 12 June 2015 at 06:30, Galen Johnson <user-87f955643e3d@xymon.invalid> wrote:
I'm pretty sure it's either...I find it a bit frustrating that we can't limit where it looks for that info...if you happen to have a non-related comment with a space or colon and a number, Xymon will attempt to parse it as a tracked variable as well. =G= From: Xymon <xymon-bounces at xymon.com> on behalf of Jeremy Ruffer < user-6d8e227afca3@xymon.invalid> Sent: Thursday, June 11, 2015 4:20 PM To: Bill Rowan Cc: xymon at xymon.com Subject: Re: [Xymon] Temperature graphs I use a colon to separate value and data. It works. JeremyOn 11 Jun 2015, at 21:03, Bill Rowan <user-3078c15a8e5e@xymon.invalid> wrote: Get rid of the spaces. Your status should read "&greenRouting_Engine_Temperature 43". Make sure you send the data in Celsius and let xymon do the conversion in the graphs.cfg file.I haven't tried using a colon for separation of data and value. I have been using just a space and it works fine for me on 4.3.7.On 6/11/2015 1:46 PM, David Fogarty wrote: I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this:/var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrdWhen I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done. Reformatting output on that may be a bit harder. The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com
--
"Accept the challenges so that you can feel the exhilaration of victory"
- General George Patton
list David Fogarty
I have removed the spaces and added a colon, so &green Routing_Engines_Temperature: 44 creates a filename that looks like: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature:.rrd And I still get no data.
▸
Dave Fogarty
--
Senior System Administrator
CollegeNET.com
On Thu, 11 Jun 2015, Galen Johnson wrote:
I'm pretty sure it's either...I find it a bit frustrating that we can't limit where it looks for that info...if you happen to have a non-related comment with a space or colon and a number, Xymon will attempt to parse it as a tracked variable as well. =G= From: Xymon <xymon-bounces at xymon.com> on behalf of Jeremy Ruffer <user-6d8e227afca3@xymon.invalid> Sent: Thursday, June 11, 2015 4:20 PM To: Bill Rowan Cc: xymon at xymon.com Subject: Re: [Xymon] Temperature graphs I use a colon to separate value and data. It works. JeremyOn 11 Jun 2015, at 21:03, Bill Rowan <user-3078c15a8e5e@xymon.invalid> wrote: Get rid of the spaces. Your status should read "&green Routing_Engine_Temperature 43". Make sure you send the data in Celsius and let xymon do the conversion in the graphs.cfg file. I haven't tried using a colon for separation of data and value. I have been using just a space and it works fine for me on 4.3.7.On 6/11/2015 1:46 PM, David Fogarty wrote: I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd When I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done. Reformatting output on that may be a bit harder. The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com
list David Fogarty
This sounds very promising. I'm assuming this would mean that I would submit the data as an alternate test name but in the rrd script, write to an rrd file named "temperature.Routing_Engine_Temperature". I'll play with this and come back.
▸
Thanks,
Dave Fogarty
--
Senior System Administrator
CollegeNET.com
On Fri, 12 Jun 2015, Vernon Everett wrote:
Hi all The simple solution is to send two messages. One, as a status, which is the display we get to see. And another with data, which we don't get to see, but gets parsed into the rrd files. Regards Vernon On 12 June 2015 at 06:30, Galen Johnson <user-87f955643e3d@xymon.invalid> wrote: I'm pretty sure it's either...I find it a bit frustrating that we can't limit where it looks for that info...if you happen to have a non-related comment with a space or colon and a number, Xymon will attempt to parse it as a tracked variable as well. =G= From: Xymon <xymon-bounces at xymon.com> on behalf of Jeremy Ruffer <user-6d8e227afca3@xymon.invalid> Sent: Thursday, June 11, 2015 4:20 PM To: Bill Rowan Cc: xymon at xymon.com Subject: Re: [Xymon] Temperature graphs I use a colon to separate value and data. It works. JeremyOn 11 Jun 2015, at 21:03, Bill Rowan <user-3078c15a8e5e@xymon.invalid> wrote: Get rid of the spaces. Your status should read "&greenRouting_Engine_Temperature 43". Make sure you send the data in Celsius and let xymon do the conversion in the graphs.cfg file.I haven't tried using a colon for separation of data and value. I have been using just a space and it works fine for me on 4.3.7.On 6/11/2015 1:46 PM, David Fogarty wrote: I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd When I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done.Reformatting output on that may be a bit harder.The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com-- "Accept the challenges so that you can feel the exhilaration of victory"- General George Patton
list Galen Johnson
Maybe I missed it in a different thread but what version are you using that still references "hobbit" paths? My rrd files don't include the colon and use a comma between test and metric. Also, is your test firing less often than every 5 minutes? My first attempt was firing off every 15 minutes and that was causing RRD to drop them since it was outside of the 10 minute default window. NCV can be a bit finicky. At that point I had 3 options: 1) check every 5 minutes (what I ended up doing) 2) use rrdtool tune after it was created to extend the default window of the rrd file 3) create the rrd myself and then let xymon populate it (which is pretty much what you have to do when you use the separate data stream). I hid my metric inside a <!-- data --> comment like devmon does and used html codes in the status area for the viewable area. Note, sometimes you have to wait up to 30 minutes before you see anything on the graph but you should be able to see values getting updated using the rrdtool dump utility.
▸
=G=
From: Xymon <xymon-bounces at xymon.com> on behalf of David Fogarty <user-ec392d47385f@xymon.invalid>
Sent: Thursday, June 11, 2015 5:28 PM
To: xymon at xymon.com
Subject: Re: [Xymon] Temperature graphs
I have removed the spaces and added a colon, so
&green Routing_Engines_Temperature: 44
creates a filename that looks like:
/var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature:.rrd
And I still get no data.
Dave Fogarty
--
Senior System Administrator
CollegeNET.com
On Thu, 11 Jun 2015, Galen Johnson wrote:
I'm pretty sure it's either...I find it a bit frustrating that we can't limit where it looks for that info...if you happen to have a non-related comment with a space or colon and a number, Xymon will attempt to parse it as a tracked variable as well. =G= From: Xymon <xymon-bounces at xymon.com> on behalf of Jeremy Ruffer <user-6d8e227afca3@xymon.invalid> Sent: Thursday, June 11, 2015 4:20 PM To: Bill Rowan Cc: xymon at xymon.com Subject: Re: [Xymon] Temperature graphs I use a colon to separate value and data. It works. JeremyOn 11 Jun 2015, at 21:03, Bill Rowan <user-3078c15a8e5e@xymon.invalid> wrote: Get rid of the spaces. Your status should read "&green Routing_Engine_Temperature 43". Make sure you send the data in Celsius and let xymon do the conversion in the graphs.cfg file. I haven't tried using a colon for separation of data and value. I have been using just a space and it works fine for me on 4.3.7.On 6/11/2015 1:46 PM, David Fogarty wrote: I've got a couple snmp tests and I'm trying to graph their output. I tried putting in my own rrd script to collect the data, but there seems to be a built-in one overriding mine. So, I tried to format my test output to work with that, but I'm just getting rrd files that lack any data. The main line of the test is like this: &green Routing Engine Temperature 43 It causes the creation of a file like this: /var/lib/hobbit/rrd/<hostname>/temperature.Routing_Engine_Temperature.rrd When I use "rrdtool dump" I still get data like this and an empty graph. <!-- 2015-06-11 11:20:00 PDT / 1434046800 --> <row><v>NaN</v></row> <!-- 2015-06-11 11:25:00 PDT / 1434047100 --> <row><v>NaN</v></row> If someone knows the correct format for temperature tests, I'd be set. Also, we've got snmp collecting memory data for some equipment that we'd like to override internal graphing on specific hosts if that can be done. Reformatting output on that may be a bit harder. The obvious answer is to change the test names for our custom ones, but my boss is very opposed to creating new columns in xymon. Thanks, Dave Fogarty -- Senior System Administrator CollegeNET.com
list Ulric Eriksson
2015-06-11 23:19 skrev Vernon Everett:
▸
Hi all The simple solution is to send two messages. One, as a status, which is the display we get to see. And another with data, which we don't get to see, but gets parsed into the rrd files.
Agreed. Sending the trend data separately is simple and makes the monitoring script cleaner. And the message format is well documented instead of a hack. Ulric