RARE user and assistance email list

[mc36 <>]

hi,
so in hbone case, we run icinga to monitor the network, and rare/freerouter 
have a server called nrpe, which once enabled,
you can easily monitor and alarm if a topology change happens... the config 
side is as easy as listing your nodes under
stanza called checks, and having the per node sho ipv4 ospf 100 topo x.x.x.x 
command... one can generate this easily
from a script once, then when a topology change happens, one just have to 
issue the knob train under the check and then
it retrieves the current, believed to be good topology, and as long as the 
network does not change, the check will
return a positive result... at the moment, frederic monitors the geant p4lab 
from bud0101 via icina, please see the check
stanzas there... similar result can be achieved if you go the prometheus way 
and export all the metrics you find in the show
commands but then you have to deal with the alerting yourself.... i 
personally find the icinga/check way more convenient,
because the network node have all the visibility and can bark back a 
meaningful message if a topology change happens...
but surely, the prometheus way is also seems doable because if a node/link 
disappears then it'll not visible in the prometheus
metric exports anymore in the sensor exports...
br,
cs


On 1/23/23 17:16, Gabriel Tetzner wrote:
> Hi Csaba,
>
> thanks for the answer, I was looking for this type of sensor for prometheus, then as I understood from your considerations, through a node of the topology, I can monitor the rest 
> of the entire topology, since ospf is a link state protocol, in which if a change is made, it is visible throughout our topology, but you said that there is not yet a sensor for 
> the command in question.
>
> Is it possible for me to create this sensor to monitor the entire ospf topology through a single node? I believe this would be my contribution to freeRouter, since I am also 
> working on monitoring it.
>
> And if it is possible to integrate twamp into it as well.
>
> One more question, is it possible to configure this ospf delay within the 
> ospf process configuration, since I've seen a command like delay metric or 
> something like that?
>
> Anyway, thanks for the answers.
>
> Sincerely,
> Gabriel Tetzner Menegueti
>
> Em seg., 23 de jan. de 2023   s 10:50, mc36 < <>> 
> escreveu:
>
>     hi,
>     niceee!
>     good observations, twamp in freerouter alters all the flexalgo metrics 
> all at once...
>     if you just use the default flexalgo 0 topology then that can profit from 
> it too...
>     regarding the prometheus, yesss its possible, as you have the show ipv4/6 
> ospf x metric knob...
>     frederic introduced similar things in the geant testbed, this is the 
> sensor he done back in the days:
>     https://github.com/rare-freertr/freeRtr/blob/master/misc/sensor/lsigp-ospf.txt 
> <https://github.com/rare-freertr/freeRtr/blob/master/misc/sensor/lsigp-ospf.txt>
>     here the perf one is that exports the it from the current node...
>     but on the other hand, please note that, as it's a link state routing 
> protocol we're talking about,
>     and the changes are reflected in ospf router lsa, that is, observable on 
> every node in the network,
>     you can monitor the whole network from a single node by the sho ipv4/6 
> ospf x topo [xxx] command...
>     imho this latter does not have yet the sensor, but should be doable too...
>     (we at hbone, we only went for the connectivity checks this way, and that 
> works fine)
>     br,
>     cs
>
>     On 1/23/23 13:52, Gabriel Tetzner wrote:
>      > Hi all,
>      >
>      > In RENaaS topology, we implemented OSPF as the internal dynamic 
> routing protocol of our infrastructure, but I wanted to know about implementing 
> the twamp service together, I
>     tried
>      > to implement twamp according to the test with ospf: 
> http://sources.freertr.org/cfg/rout-ospf57.tst 
> <http://sources.freertr.org/cfg/rout-ospf57.tst>
>     <http://sources.freertr.org/cfg/rout-ospf57.tst 
> <http://sources.freertr.org/cfg/rout-ospf57.tst>> , but I noticed that twamp in
>      > this case helps in managing the cost of ospf, am I right?
>      >
>      > I say this because I want to do the OSPF link delay configuration:
>      > 
> https://www.juniper.net/documentation/us/en/software/junos/ospf/topics/topic-map/enable-link-delay-advertise-in-ospf.html
>     
> <https://www.juniper.net/documentation/us/en/software/junos/ospf/topics/topic-map/enable-link-delay-advertise-in-ospf.html>
>      > 
> <https://www.juniper.net/documentation/us/en/software/junos/ospf/topics/topic-map/enable-link-delay-advertise-in-ospf.html
>     
> <https://www.juniper.net/documentation/us/en/software/junos/ospf/topics/topic-map/enable-link-delay-advertise-in-ospf.html>>
>    and do the network performance verification with
>     twamp,
>      > I think this is possible in freeRouter. I wonder if there are other 
> tests with OSPF and twamp that I can do to measure network performance and if it 
> is possible to integrate
>     this
>      > with prometheus and grafana?
>      >
>      > Yours truly,
>      > Gabriel Tetzner Menegueti 
>
> _._,_._,_
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> Groups.io Links:
>
> You receive all messages sent to this group.
>
> View/Reply Online (#992) <https://groups.io/g/freertr/message/992> | Reply To Group 
> <> | Reply To Sender 
> <> | Mute This Topic 
> <https://groups.io/mt/96472567/6006518> | New Topic <https://groups.io/g/freertr/post>
> Your Subscription <https://groups.io/g/freertr/editsub/6006518> | Contact Group Owner 
> <> | Unsubscribe <https://groups.io/g/freertr/unsub> 
> []
>
> _._,_._,_