Re: [zenoss-users] Events not being created after i make some TIGHTthresholds

Wed, 21 Feb 2007 07:59:39 -0800 

Hi Daniel,
You should get threshold events, if you drop a threshold below the current value. There's nothing tricky going on. It may take a while to propagate the change to the collector, since the collector only re-reads its configuration every 20 minutes by default. 


Check your log for the collector (probably zenperfsnmp) to make sure 
that events are being sent.


Try running the collector against your device in verbose mode:

   $ zenperfsnmp run -v 10 -d my.device.name --cycle

-Eric

Daniel McKinney wrote:

Kristopher,


No luck.  :(  I set a threshold outside of normal limits then did an 
snmpwalk on another server to test the outgoing threshold.  I also 
increased the incoming threshold as well and both thresholds were 
crossed.  No events were created and nothing in the logfiles that i 
can find.  Kristopher, could you replicate this reliably?  You can set 
a threshold below the current data levels and it wont generate an 
alert, but create a threshold outside of normal data levels, then 
cross the lines and it will generate alerts?



Just as an FYI, other events are being generated for the same server 
and other servers, just not the NIC Utilization thresholds on this or 
any servers.


Thanks,

-Daniel

Daniel McKinney wrote:

Kristopher,

Thanks for the quick reply.


I will defiantly try this to see if that will at least create an 
event.  One thing to note though is that when i do the same process 
on the VMware Zenoss, it will generate events even though the initial 
data is past that threshold.  So i am wondering if there is some 
setting that would cause this behavior.


I will follow up with the results of your suggestion.

Thanks much,

-Daniel

Kristopher D. Dick wrote:

I've noticed similar behavior on other things.  It appears that 
threshold

events get triggered when the threshold is crossed, so that if you're
already above the threshold when the threshold is defined, it doesn't
trigger the threshold event.  I may be wrong, but that's what I've

experienced.  Try setting a higher threshold, and then put an 
absurdly large

load on it.

Kristopher Dick

-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of switchfox
Sent: Tuesday, 20 February 2007 09:32
To: [email protected]
Subject: [zenoss-users] Events not being created after i make some
TIGHTthresholds

Hello all,


I wonder if anyone has come across this.  I have seen similar emails 
but

nothing exactly like this and it is very frustrating.

I am trying to set thresholds on any server in my network for port
utilization of the Throughput.

By default, Zenoss gets installed with a "Utilization_75_perc" 
threshold
template that looks at the ifInOctets and ifOutOctets data sources.  
I look
inside that RRD Threshold and the Max Value field has the following 
in it:

"here.speed / 8 * .75".


This threshold is disabled by default, which is fine.  I wanted to 
see if i
could create two new threshold checks for both out or in traffic 
separately.

So i created two new Thresholds at the Global Devices PerfConf:

http://172.16.10.143:8080/zport/dmd/Devices/rrdTemplates/ethernetCsmacd 
and

called them "Utilization_1_perc_IN" and "Utilization_1_perc_OUT".


Seeing as i wanted to create lots of noise for testing, i set the 
threshold
to insanely tight as the following shows (Also please see the 
attached graph

to show the math actually worked out to show that the first and second
thresholds are both set to 1 kilobit):
Utilization_1_perc_IN
Max Value = here.speed / 8 * .000001

Utilization_1_perc_OUT
Max Value = here.speed / 8 * .000001


Now, seeing that the thresholds are really tight and ANY traffic 
should send

out an alert, i would think that if i looked at the Events tab for ANY
device on my network, i should be seeing long lists of
"monitor1.sunrocket.com Utilization_1_perc_OUT ..." events.

It has been a while since i added these thresholds and still no 
event has
been sent.  I tried bouncing the zenxevent daemon thinking since it 
was a

configuration change that it might need to be bounced to pick up that

change.  I even bounced all of Zenoss to see if that was needed, 
still no

event.


It is important to note that if I do this entire process on the 
VMware image
that Zenoss provides (1.0 or 1.1) it works fine, i get alerts all 
day long.



What setting should i check, or what logfile should i begin looking 
at to

further troubleshoot?  I see alot of entries for Utilization_TEST... in
Z2.log, but i am assuming that is just the http access log for Zenoss.

Any help would be greatly appreciated,

-Daniel






_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users














    











					Reply via email to