More Infos on Wireless Link quality
As we’ve all have wireless problems from time to time we need
every bit of information which helps us. So I would like to see
a counters for packet errors/retransmission, CCQ for each link. And
when these values are graphed I can see cyclic problems.
Yes I like ROS very much, but I have some Proxim Tsunamis and
Cisco Aironets out there. And I would like to see signal strength for
them too.
Wireless Client Table.
I do not want to see all wireless clients on an accesspoint graphed
permanently, but if I touch the AP Icon and there comes up a table
with all clients of this AP and there signal strength …
Colors of Links. If graphs grow its hard to see when there are
lines near there capacity limit. So I would like to see them yellow
on half speed and orange if they are near the performance limit.
Red if there is a problem (signal strength below a limit/retry count
above a limit).
we could not know what your wireless link capacity is. maybe we could make some condition->action functionality. for example if $signal>-40 then $link_color=red
Yes. This would do it. If I can write conditions like this:
$signal>-40 then $link_color=red
$retries>100 then $link_state=blink
$bps>10000 then $link_color=yellow
At the moment I use a monitoring tool that does the color job
and uses “ants” (small points/lines flowing along the link) to
show busy lines. There are
no ants when no traffic at all
small ants when there is traffic
big ants when there is a lot of traffic
yellow link when line gets realy busy
orange link when line is full
So I can get a feeling what my lines are doing very quick
The solution I’m using has some drawbacks (not only the price )
so I look into using dude.
I’m thinking about a sort of bandwidth functionality within The Dude.
In the environment I’m working in we also have about 20 environmental probes from ACKP for temperature and humidity monitoring. Besides these we have 2 different UPS’s, both from Eaton/Powerware.
All these objects are SNMP enabled and what we would like to be able to monitor seems to be a hell of a job to implement but we think it might be a usefull feature..
For both environmental probes as UPS’s it might be convenient if you would be able to monitor the values of given OID’s, and change the state of the device to a colour/condition defined by its value.
To be more exact..
Condition: Values (examples !): Colour:
HIGH CRITICAL > 30 or higher Red
HIGH WARNING > 28, 29 Orange
HIGH ALERT > 26, 27 Yellow
NORMAL (OK) > 20 - 25 Green
LOW ALERT > 18, 19 Yellow
LOW WARNING > 16, 17 Orange
LOW CRITICAL > 15 or less Red
I hope I’m clear enough.. If not feel free to drop me a note and I’ll try to explain better what on Earth I’m trying to explain..
BTW: it might be very usefull if the returned values from the OID’s would be graphed (seperately or joined in one major graph). It’s not that usefull to monitor something with SNMP and not being able to graph the returned values.. (that’s our experience anyway.. )
You can graph all you want right now… for example, I’m graphing the running of particular processes on some Windows machines (for example, graph is 1 when process is running, 0 if not). Just create a probe for it and apply that probe to the devices you want.
Well, if it is possible to graph the regularly changing OID-value for the temperature/humidity measured with a AKCP environmental probe or Voltage/Amperage/Load on a UPS. Please be so kind to shed some light on the how-to do that..
Until now I haven’t been able to work something like this out, all I get in the graphs are the responstimes (also in the legend) and that’s something different than the actual values don’t you think? I did manage to watch the status of the probes 2 = normal, others are faulty, but what I really want to see isn’t the status but the value of the OID.. Such as for humidity 45(%), not 2, for temperature 22 (C) not 2.. For humidity 75(%) not 3 etc..
Hope this made a bit clearer what I actually mean and want, maybe you (or anybody else) can point me out how to do this? Any help is greatly appreciated..
After a little struggle I managed to get graphs from the values I want to monitor. I’ve to say it’s not that easy to figure it all out, but thanks to your example a human has got to be very stupid if you won’t be able figure it out someday..
we need