New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Self Hosted Monitoring for my needs
Hi Guys,
Looking for some suggestions here. The debate of which is best monitoring solution can be lengthy so I will get straight to my requirements.
I need to monitor 5-10 servers (My own cloud servers, have full access)
And I need to monitor about 100 sites. (Not mine but for a community project hosted with different providers. Have no server access)
So is there any Open source tool that can be used? Need to keep it simple but workable.
Must be able to have graphs (for uptime, response time for sites) and server resource usage.
Also should be able to use multiple locations for monitoring. (i.e. 3-5 locations)
Thanks
Umair
Comments
i think 0xdragon just launched something , i might be mistaken.
Yep, I run a hosted monitoring service. But I think @Umair wants a self hosting one. However, at the price of $4/yr for your own install and unlimited usage, it might be worth considering my service.
CloudCommando Hosted Observium @ $4/year
Or try CloudCommando hosted Observium for one month for only $1!
I was looking into this recently and Munin looked promising to me. edit: It provides many graphs right out of the box IIRC.
http://munin-monitoring.org/
seriously .. how can you go wrong with monitoring for 4 bucks a year
Are you looking for uptime monitoring or more complex monitoring (like memory/CPU usage)? You can always use uptime monitoring to accomplish the latter but that might not be optimal.
Anyway, for a very simple distributed uptime monitoring scheme, see https://github.com/uakfdotb/pybearmon (it is designed to provide minimal functionality to get a reliable monitoring system from multiple locations).
This doesn't do the graph generation for you, but it does record historical uptime data that you can create graph from.
If you are looking for something more fancy and feature-rich consider nagios. I don't think observium supports distributed monitoring?
Apologize to hijack thread, but will i be able to add extra users in $4 plan ? Like if i want someone else to see my servers as well ?
As per the CloudCommando website:
It's your own isolated instance of Observium, so feel free to do whatever you want with it! :-)
(So long as you follow the rules.)
Thank you for the explanation
Couldn't see much information on old design, new one is awesome.
Hello,
I am a little lost i thought Umair was seeking info to make his own monitoring server.
If i am wrong i am sorry, the first names that rings a bell are cacti and nagios.
As a piece of advice you can also ask your provider/s if they do not offer this service
It might be limited to icmp but it might be enough.
I use https://github.com/fzaninotto/uptime Runs on node.js and mongodb, it shows uptime, response time and alerts.
I use this for when i want to monitor if the site is up or down. But it depends if he needs cpu and other stats monitoring.
Monit has some good features such as monitoring services and restarting them if need be.
https://github.com/Munroenet/serverstatus
@0xdragon
I assume that for the huge sum of $4 there's no support expectation but that said, I've sent you a reply on the "deploy@" mail with some Qs
But I don't know if that goes on an inbox or nowhere at all. Please let me know if you've received it or not (no need to answer it, just wondering).
Ah, it should've been forwarded to my inbox. Might've hit spam! I'll get to your query later today, I'm incredibly busy with migrations etc. Apologies for not answering sooner!
Thanks!
Migrations... Hmmmm,
I'm waiting for my observium
When will I be able to log in again?
Mate, I can appreciate that you're impatient to add more servers, but as I've made very clear, I'm a one man team trying to work on this huge migration which is literally changing the entire platform. You don't need to open support tickets and ping me on here at the same time.. I have to get this done.
Please let me know if you have any further queries.
Thanks,
Alexander.
You are not lost
You are absolutely right. I can use multiple services for monitoring my site/server. But for that "community project" I do need to have a self hosted solution. cacti and nagios are pretty good but I think they are a lil too much for my needs. (And complex to setup initially).
I mean, I only have 50-70 sites to monitor. (Site's not server)
I do need to be able to monitor from multiple locations (3-5)... (to confirm site is indeed down). Need graphs for response time etc as well.
@Umair Ok, so my advice remains to ask your providers if they have an icmp monitoring service if it is enough for you
+1 for CloudCommando. @0xdragon
does any of there services measure daily bandwidth used ? I'd like to have a weekly report sent to me, per domain. I'm using VestaCP & Webmin on different VPS
I've been running Zabbix recently and quite like it, although thats more for resource/bandwidth monitoring rather than uptime monitoring and mine's only monitoring a few hosts although it will fire off an alert if it loses contact with one of the boxes it's monitoring.
Got it running on one of those 2euro dedis.
I also use uptime roboot for monitoring if certian sites/boxes are up.
Icinga2 is also in this area. Best part is that you can monitor from multiple places and have it report into a central database. As long as your servers respond from one location, no inbox or SMS meltdown
@0xdragon can help with hosting Icinga2 also. You can add extra tiny VPS's in various locations to report into it which kind of helps in the 'self hosted' stakes. You would even run a small instance at home to ping out too.
Observium won't work if you don't have root server acces, how about this though?
https://github.com/uakfdotb/pybearmon
Huh?
Ah, makes sense. I didn't interpret your comment the correct way
No root!!
I'm also a psychic :O