RARE user and assistance email list

[Frédéric LOUI <>]

Hi !

Let me give you the approach here:
1- Check the KPI is available and viewable from freeRtr CLI. 
    Sometimes the KPI is here but the CLI output has format difficult to 
parse from Prometheus. 
    In that case the RARE team can adjust the output in tabular format so 
that it is easy to be read by prometheus or influxdb/telegraf
    —> this is where a small development is need from our part

2- Create a specific sensor object defining the CLI command and the parsing 
method 

3- Create: 
    b- prometheus server 
    b-or telemetry exporter

Enjoy, data collected on prometheus db or influxdb via telegraf.

Now you have the raw data you can write a small agent script that would feed 
Timemap.
If TImemap is using grafana, you just have to add the prometheus/influx as 
new grafana sourcesThen the next step is to adjust your grafana dashboard 
with the newly added source.

If Timemap has its own database, then you’d need to write a small agent 
script that would fetch data from prom/influx and make it into TimeMap db 

All in all this can be done quickly for monitoring expert familiar to 
Grafana/Influx/Prometheus stack

All the best,
Frederic

> Le 23 août 2022 à 17:01, mc36 <> a écrit :
> 
> 
> 
> On 8/23/22 16:51, Tim Chown wrote:
>> On 23 Aug 2022, at 15:46, mc36 <> wrote:
>>> 
>>> well, we all know what jitter is, rtt max - rtt min, also in my ping...
>> Or the average deviation of delays from the average delay.
> yeahhh, so these formulas normally should be interchangeable,
> that is, if you calculate the stuff in the one way then you should arrive 
> to the same result to the other way...
> 
> i personally used the mean deviation for the ping because in such a low 
> intervals it should not make any difference,
> and most notably, it's very easily to calculate from a series of values 
> without having to store the whole dataset:
> 
> https://en.wikipedia.org/wiki/Standard_deviation#Rapid_calculation_methods
> 
> the other formulas, well... would require some more memory and cpu to have 
> on the routers, but doable.... :)
> 
> 
> 
> 
>>> but still, once you have the series of rtt, imho all the others in the 
>>> timemap just are a matter of calculation,
>>> should be done elsewhere not on the router, or probably it's already 
>>> doable in the promsql... frederic!?
>> Indeed, only the delays are gathered, the rest is calculated off-router.
>> Tim
>>> 
>>> 
>>> On 8/23/22 16:40, Tim Chown wrote:
>>>> Hi,
>>>> Here’s a deck from March 2022 about the components used in TimeMap, 
>>>> slide 8 I think is most useful:
>>>> https://events.geant.org/event/1084/contributions/1022/attachments/672/942/Timemap_Anomaly_Detection_2ndPMW_2022_FFv1.pdf
>>>> Calculation of jitter can be a subtle, nuanced thing.
>>>> Tim
>>>>> On 23 Aug 2022, at 15:33, mc36 <> wrote:
>>>>> 
>>>>> 
>>>>> 
>>>>> On 8/23/22 16:23, mc36 wrote:
>>>>>> On 8/23/22 16:18, mc36 wrote:
>>>>>>> well, doable, and it's a separate class if i recall correctly, that 
>>>>>>> is, readily reuseable,
>>>>>>> and really not too time comsuming O(n), and i already do so for the 
>>>>>>> ping because at that point,
>>>>>>> it's useful to have at hand, but heyy, once you have the stream, why 
>>>>>>> would one do so???
>>>>>> lemme reverb this real quick...
>>>>>> so that it's a separate class in freerouter, that is reuseable, but,
>>>>>> since it's a calculated value from a series of data,
>>>>> 
>>>>> so let's say we're already exporting the rtt as time series data,
>>>>> then to have the deviation from the series is just a matter of 
>>>>> calculation...
>>>>> you can do it on the router surely, but if you do it elsewhere then you
>>>>> must be able to select a time range to calculate the deviation from 
>>>>> just those...
>>>>> 
>>>>> the thing is that, it could be a completely different value:
>>>>> 
>>>>> please take a look at the below two pings... i tried to start them at 
>>>>> the same time...
>>>>> they calculate the deviation obviously the same way, the difference is 
>>>>> the time interval,
>>>>> 1111 pings: min/avg/max/dev rtt=3/4.0/19/1.0
>>>>> 2222 pings: min/avg/max/dev rtt=3/4.0/19/1.2
>>>>> 
>>>>> probably, i had some more bigger responses when it ran twice the time 
>>>>> and that's why the mean deviation 1.0 vs 1.2 mean deviation...
>>>>> 
>>>>> br,
>>>>> cs
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> noti#ping www.niif.hu repeat 1111
>>>>> info userReader.cmdEnter:userReader.java:1087 command noti#ping 
>>>>> www.niif.hu repeat 1111  from local:telnet <loop> 23 -> 127.0.0.1 58682
>>>>> 2022-08-23 16:31:17
>>>>> resolving www.niif.hu for ipv6 ok!
>>>>> pinging 2001:738:0:420::b, src=2001:db8:1101::11, vrf=inet, cnt=1111, 
>>>>> len=64, df=false, tim=1000, gap=0, ttl=255, tos=0, sgt=0, flow=0, 
>>>>> fill=0, alrt=-1, sweep=false, multi=false
>>>>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!info
>>>>>  userReader.cmdEnter:userReader.java:1087 command noti#ping www.niif.hu 
>>>>> repeat 2222  from local:telnet <loop> 23 -> 127.0.0.1 35006
>>>>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>>>>> result=100.0%, recv/sent/lost/err=1111/1111/0/0, took 4603, 
>>>>> min/avg/max/dev rtt=3/4.0/19/1.0, ttl 51/51.0/51/0.0, tos 0/0.0/0/0.0
>>>>> noti#
>>>>> 
>>>>> 
>>>>> noti#ping www.niif.hu repeat 2222
>>>>> info userReader.cmdEnter:userReader.java:1087 command noti#ping 
>>>>> www.niif.hu repeat 2222  from local:telnet <loop> 23 -> 127.0.0.1 35006
>>>>> 2022-08-23 16:31:17
>>>>> resolving www.niif.hu for ipv6 ok!
>>>>> pinging 2001:738:0:420::b, src=2001:db8:1101::11, vrf=inet, cnt=2222, 
>>>>> len=64, df=false, tim=1000, gap=0, ttl=255, tos=0, sgt=0, flow=0, 
>>>>> fill=0, alrt=-1, sweep=false, multi=false
>>>>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>>>>> result=100.0%, recv/sent/lost/err=2222/2222/0/0, took 9184, 
>>>>> min/avg/max/dev rtt=3/4.0/19/1.2, ttl 51/51.0/51/0.0, tos 0/0.0/0/0.0
>>>>> noti#
>>>>>