[hobbit] Graphing with broken numbers
----- "thorsten erdmann" <thorsten.erdmann at daimler.com> wrote:
Hi
I try to get some temperature graphs of my UPS. The UPS test produces broken numbers like this:
What do you mean by "broken"? Do you mean non-integers?
Temperature status:
Warning level: 35 C Alarm level : 40 C
Device Temp(C) Temp(F)
Ambient 20.7 69
Devmon version 0.3.1-beta1 running on s068a300
I manged to get the numbers into a graph if I alter the output to not showing 20.7, instead showing 21. The graph show up several lines like
Ambient_______20. Ambient_______21.
I am not sure what you are trying to depict here, are you meaning that the DS names are Ambient______20 ? The output of 'rrdtool info /path/to/temp.rrd |grep ^ds' or similar may help to illustrate more accurately.
Which collector are you using?
Regards, Buchan
First, sorry for the late response, I was on vacation.
What do you mean by "broken"? Do you mean non-integers? Yes, I mean non integer, e.g.: 22.3 deg or in german notation 22,3 °C
I am not sure what you are trying to depict here, are you meaning that the DS names are Ambient______20 ? The output of 'rrdtool info /path/to/temp.rrd |grep ^ds' or similar may help to illustrate more accurately. I mean, that the rrd graphs show up a line for each integer that ever has occured. e.g. if I have an output line in the temperture test of
"Ambient 20.5"
I get a line named "Ambient_____20." in the rrd graph with the value 5. So rrd recognizes the string up to the colon as the name and the part after the colon as the value. So if now the temperature changes to 21.2 I get a new line in the rrd graph:
Ambient_____20. with value 5 Ambient_____21. with value 2
Which collector are you using? I have written the test by myself and it produces an output like the standard temperature test. I therefore use the standard temperature collector.
This output works:
Thu Oct 1 08:30:59 CEST 2009 Temperature Test on webtherm1
Warning level: 28 C Alarm level : 30 C
Device Temp(C) Temp(F)
green Webtherm1 24 75
This output does not work:
Thu Oct 1 08:30:59 CEST 2009 Temperature Test on webtherm1
Warning level: 28 C Alarm level : 30 C
Device Temp(C) Temp(F)
green Webtherm1 24.5 75.5
If you are not the intended addressee, please inform us immediately that you have received this e-mail in error, and delete it. We thank you for your cooperation.
Hi
I have just noticed that devmon stops working if a device it should test is not reachable. We had an UPS test running via devmon an one of the UPS is uninstalled but not deleted in Hobbit/devmon. When I start devmon it does exactly one scan which reported tons of errors in the devmon log and then nothing happens any more. The Hobbit devmon tests of all devices get purple after some time.
How can I setup devmon to ignore such errors and not stop working.
Thanks Thorsten
If you are not the intended addressee, please inform us immediately that you have received this e-mail in error, and delete it. We thank you for your cooperation.
On Thursday, 1 October 2009 07:47:04 thorsten.erdmann at daimler.com wrote:
Hi
I have just noticed that devmon stops working if a device it should test is not reachable. We had an UPS test running via devmon an one of the UPS is uninstalled but not deleted in Hobbit/devmon. When I start devmon it does exactly one scan which reported tons of errors in the devmon log and then nothing happens any more. The Hobbit devmon tests of all devices get purple after some time.
One of our UPS's was the last to be moved to a new network, and this caused no problems.
Can you provide more detail: 1)What version of devmon are you running? 2)Is hobbit still testing this device? 3)Was there a conn test for this device? Is it still there? What colour is it currently 4)Can you provide the useful parts of the log?
How can I setup devmon to ignore such errors and not stop working.
Usually devmon should not poll the device if the conn test is red, and even if it does, it should at worst time out for the maximum of (SNMPTIMEOUT * SNMPTRIES) and MAXPOLLTIME. While it is timing out the fork that is polling this device, the other forks should continue polling ....
Regards, Buchan
participants (2)
-
bgmilne＠staff.telkomsa.net
-
thorsten.erdmann＠daimler.com