(my interpretation of it)
The colours are just an easy way to see which devices are taking longer to discover/poll.
The percentage is based around the longest individual device time (that is currently up/responding). The slowest active device gets 100% and all of the others are scored based on their fractional time taken.
I have a device that is down with a poller of 139%.
It helps to identify if you could do with more pollers (ie. you have more devices red/orange than you have poller threads), or if you just have one or two rogue devices that are taking excessively long to process (in which case, perhaps consider disabling some modules/mibs for that device).
Don't forget to keep in mind that you want the complete poller cycle to complete in under 5 minutes. I wouldn't bother trying to tune rogue devices unless they are creeping up on that 5 minute mark. It can also be somewhat beneficial (if a little counter-intuitive) to "slow down" the poller cycle so that it is near 5 minutes for a complete run; this reduces the overall load on your Observium box and means it won't go unresponsive trying to deal with huge spikes every polling interval.
Hope this helps,
Michael
On 19 Feb 2017, at 8:24 pm, Youssef BENGELLOUN - ZAHR ybzahr@prodware.fr wrote:
Dear Observium community,
Can someone shed some light on the meaning of color coding used in the poller page please ?
<image001.png>
I can gather on my own that colors seem to indicate some kind of performance result : GREEN is good, BLUE or PURPLE is average, RED is bad. I don’t understand what the percentage figures seem to indicate.
Are those in any way related to poller performance issues on the NMS server ? meaning, does that imply that I should maybe increase NMS server ressources (CPU, memory, etc.) ?
Thank you.
Youssef BENGELLOUN - ZAHR - Consultant Expert Prodware France T : +33 979 999 000 - F : +33 988 814 001 - ybzahr@prodware.fr <image325336.PNG> Web : prodware.fr
<image437576.JPG> <imagec49228.JPG> <imaged44e86.JPG> <imageeee5e3.JPG> <imagef6e000.JPG> <image577b3e.JPG>
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Dear Michael,
That was, more or less, my understanding as well.
Thank you for your feedbacks.
Best regards.
Le 19/02/2017 12:13, « observium au nom de Michael » <observium-bounces@observium.org au nom de obslist@smarsz.com> a écrit :
(my interpretation of it)
The colours are just an easy way to see which devices are taking longer to discover/poll.
The percentage is based around the longest individual device time (that is currently up/responding). The slowest active device gets 100% and all of the others are scored based on their fractional time taken.
I have a device that is down with a poller of 139%.
It helps to identify if you could do with more pollers (ie. you have more devices red/orange than you have poller threads), or if you just have one or two rogue devices that are taking excessively long to process (in which case, perhaps consider disabling some modules/mibs for that device).
Don't forget to keep in mind that you want the complete poller cycle to complete in under 5 minutes. I wouldn't bother trying to tune rogue devices unless they are creeping up on that 5 minute mark. It can also be somewhat beneficial (if a little counter-intuitive) to "slow down" the poller cycle so that it is near 5 minutes for a complete run; this reduces the overall load on your Observium box and means it won't go unresponsive trying to deal with huge spikes every polling interval.
Hope this helps,
Michael
> On 19 Feb 2017, at 8:24 pm, Youssef BENGELLOUN - ZAHR ybzahr@prodware.fr wrote: > > > Dear Observium community, > > Can someone shed some light on the meaning of color coding used in the poller page please ? > > <image001.png> > > I can gather on my own that colors seem to indicate some kind of performance result : GREEN is good, BLUE or PURPLE is average, RED is bad. I don’t understand what the percentage figures seem to indicate. > > Are those in any way related to poller performance issues on the NMS server ? meaning, does that imply that I should maybe increase NMS server ressources (CPU, memory, etc.) ? > > Thank you. > > > Youssef BENGELLOUN - ZAHR - Consultant Expert > Prodware France > T : +33 979 999 000 - F : +33 988 814 001 - ybzahr@prodware.fr > <image325336.PNG>Web : prodware.fr > > <image437576.JPG> <imagec49228.JPG> <imaged44e86.JPG> <imageeee5e3.JPG> <imagef6e000.JPG> <image577b3e.JPG> > > > >
BENGELLOUN - ZAHR Youssef - Consultant Expert Prodware France T : +33 979 999 000 F : +33 988 814 001 - ybzahr@prodware.fr Web : prodware.fr
_______________________________________________ > observium mailing list > observium@observium.org > http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
_______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
Yes. Our recommendation is to aim for as close to 300 seconds as possible for the poller duration, that way you're not creating periods of high and low load slowing down the web ui.
Adam.
Sent from BlueMail
On 19 Feb 2017, 11:13, at 11:13, Michael obslist@smarsz.com wrote:
(my interpretation of it)
The colours are just an easy way to see which devices are taking longer to discover/poll.
The percentage is based around the longest individual device time (that is currently up/responding). The slowest active device gets 100% and all of the others are scored based on their fractional time taken.
I have a device that is down with a poller of 139%.
It helps to identify if you could do with more pollers (ie. you have more devices red/orange than you have poller threads), or if you just have one or two rogue devices that are taking excessively long to process (in which case, perhaps consider disabling some modules/mibs for that device).
Don't forget to keep in mind that you want the complete poller cycle to complete in under 5 minutes. I wouldn't bother trying to tune rogue devices unless they are creeping up on that 5 minute mark. It can also be somewhat beneficial (if a little counter-intuitive) to "slow down" the poller cycle so that it is near 5 minutes for a complete run; this reduces the overall load on your Observium box and means it won't go unresponsive trying to deal with huge spikes every polling interval.
Hope this helps,
Michael
On 19 Feb 2017, at 8:24 pm, Youssef BENGELLOUN - ZAHR
ybzahr@prodware.fr wrote:
Dear Observium community,
Can someone shed some light on the meaning of color coding used in
the poller page please ?
<image001.png>
I can gather on my own that colors seem to indicate some kind of
performance result : GREEN is good, BLUE or PURPLE is average, RED is bad. I don’t understand what the percentage figures seem to indicate.
Are those in any way related to poller performance issues on the NMS
server ? meaning, does that imply that I should maybe increase NMS server ressources (CPU, memory, etc.) ?
Thank you.
Youssef BENGELLOUN - ZAHR - Consultant Expert Prodware France T : +33 979 999 000 - F : +33 988 814 001 - ybzahr@prodware.fr <image325336.PNG> Web : prodware.fr
<image437576.JPG> <imagec49228.JPG> <imaged44e86.JPG>
<imageeee5e3.JPG> <imagef6e000.JPG> <image577b3e.JPG>
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
These colours are just the observium standard percentage colours as used elsewhere in the ui.
They are relative to the slowest polling device, iirc. They're just a vague indicator and have no specific meaning as such.
Adam.
Sent from BlueMail
On 19 Feb 2017, 09:24, at 09:24, Youssef BENGELLOUN - ZAHR ybzahr@prodware.fr wrote:
Dear Observium community,
Can someone shed some light on the meaning of color coding used in the poller page please ?
I can gather on my own that colors seem to indicate some kind of performance result : GREEN is good, BLUE or PURPLE is average, RED is bad. I don’t understand what the percentage figures seem to indicate.
Are those in any way related to poller performance issues on the NMS server ? meaning, does that imply that I should maybe increase NMS server ressources (CPU, memory, etc.) ?
Thank you.
Youssef BENGELLOUN - ZAHR - Consultant Expert Prodware France T : +33 979 999 000 - F : +33 988 814 001 - ybzahr@prodware.fr _____________________________________________ Web : prodware.fr
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
participants (3)
-
Adam Armstrong -
Michael -
Youssef BENGELLOUN - ZAHR