[sdnog] another way of visualising your stats.

[Samir Abdullatif]

I’ve had good experience with InfluxDB. it integrates well with Grafana. And has their own SNMP Collector Telegraf which is not very complicated to setup. Downside is that the free version does not support high availability.

One point to consider, is that you need to plan carefully for your retention policies (i.e. how long you are planning to store the data). Unlike Cacti or other RRD-based monitoring systems, storage can be expensive with time-series DBs especially in Sudan where less access to commodity hardware.

I’d suggest using time-series DB for shorter retentions with more granularity and then either “Downsampling” for longer trends or use other RRD tool with bigger polling intervals.

If you already have cacti in place there is a plugin (CereusTransporter) that allows you to automatically export the polled data from Cacti directly into an InfluxDB, Bosun or OpenTSDB. This will allow you to avoid the hassle of dealing with all the SNMP OID/MIB configuration and configure your devices from the Cacti UI and then have the data available in InfluxDB and Grafana.

-
Samir

From: Philip Paeps<mailto:philip at trouble.is>
Sent: Thursday, 3 February 2022 5:36 AM
To: Nishal Goburdhan<mailto:nishal at controlfreak.co.za>
Cc: Sudan NOG<mailto:sdnog at sdnog.sd>
Subject: Re: [sdnog] another way of visualising your stats.

On 2022-02-02 20:03:47 (+0800), Nishal Goburdhan wrote:
> i know from questions i’ve been asked at sdnog, that stats/nm
> systems are an area of interest for a lot of people.  everybody knows
> about cacti, and i recall many chats about librenms.  well, here’s a
> very recent post on how you can use grafana to accomplish the same,
> using slightly more modern tools.
> https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgrafana.com%2Fblog%2F2022%2F02%2F01%2Fan-advanced-guide-to-network-monitoring-with-grafana-and-prometheus%2F&data=04%7C01%7C%7C5e1cfb763cac4dcf0cb108d9e6b58f8b%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637794489761100947%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=nY0nUeDgMuBhythz1qSzPPYM%2FFvtYGnXNcir4hvRDes%3D&reserved=0
>
> time-series captures are no longer difficult to setup;  and they can
> help to expose situations that other tools might miss.  so if NMS is
> your area of interest, this should be useful to you.

I can also recommend Prometheus / AlertManager as a slightly less
aggravating alternative to Nagios.  If pretty Grafana stats aren't
attractive enough. ;)

(But all alerting systems suck.  Systems shouldn't ever fail.  Or they
should at least be polite enough to fail when you've got time for them
and are looking at them anyway.)

Philip

--
Philip Paeps
Senior Reality Engineer
Alternative Enterprises
_______________________________________________
sdnog mailing list
sdnog at sdnog.sd
https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.sdnog.sd%2Fmailman%2Flistinfo%2Fsdnog&data=04%7C01%7C%7C5e1cfb763cac4dcf0cb108d9e6b58f8b%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C637794489761100947%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=b2dvH9grgrCdNZRy5TDMu2UxnLvTbYxglzpDtEIEhY0%3D&reserved=0

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sdnog.sd/pipermail/sdnog/attachments/20220203/2b596c72/attachment.htm>