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.
Comments
You are planning to provide this as a service or make it self hosted? In first case maybe you can group monitored IPs by DC or location and in case of down time you can report that N number of users with this DC/providers reports downtime too
An api / interface to update DNS records as actions would be awesome. As mentioned custom commands on regular intervals or as action would be extremely useful. For example checking the mail queue on regular intervals will give an idea if the server is compromised to send spam.
Adding a SLA monitor / service credit calculator would help gauge reliability of the service
Once the server is back up after a downtime, running uptime or similar commands to identify the cause ( node down / network glitch / ddos etc., ) will be more meaningful rather than up down logs.
On the network side, I monitor my IP for their ASN, ISP, organization, rdns etc., so if you can integrate such features as seen fit would make your service stand out with the regular ping checkers...
SNMP polling or trap destination?
Thanks for the suggestions guys!
@jimpop Grouping is an interesting idea and is something I can add no problem.
@Makkesk8 1 minute checks, sure. But is there really any significant need or interest for 10-30 second checks? It would need some changes to the architecture but possible as a premium feature if there's a market.
@mikecupcake The focus is a website and API right now. I could easily create a script that connects to the API and outputs/updates details if needed.
@Sentinel There will be a RESTful API.
@gbshouse This will be a service, and that's a good idea, thanks! It'd be useful to detect situations in which a whole DC is down.
@peppr Updating DNS records is an action I plan on adding, but it will need to support certain providers. Connecting via SSH and running a command and checking the response is also a possibility. Another (premium) possibility considered was allowing Python scripts to be added and run on the server (Python as most/all Linux ships with it).
The SLA monitor feature you mean for the service? I have something like this in mind but with automated payouts in false alert or failed downtime detection. Nothing is 100% on it yet and I'm still tweaking the way it works.
Again connecting to the server and running commands isn't a problem. But I don't want to go down the route of needing users to install anything on the server/device.
By monitoring, you mean checking if any of these changes? I can add a whois and notify if changed option easily. If something more complex please let me know.
@Zshen SNMP polling outbound from the service? I had assumed most people restricted SNMP from access from the open Internet.
--Adam
Maybe if the server is down it can provide an MTR or Traceroute allowing a user to better troubleshoot an issue.
@mcmyhost This is great for when a down event is detected. However, at least initially, troubleshooting the downtime isn't something that is amenable to automation on a wide scale.
Things such as if SMTP/port 25 down, then connect to server and check SMTPd is running and if not start/restart are doable as the problem can be easily identified. The SMTPd is either running or it isn't.
For a system to accurately determine at which point in the network a problem is occurring and then to intelligently do something about it is a significantly harder problem. It is something I'd really like to do, but for launch isn't something I can see being available in an advanced state.
--Adam
@adly I would think opening up to a known host wouldn't be an issue? Unless the monitoring will be coming from unknown and randomly switching IP's, this shouldn't be an issue.
@Zshen Fair enough, I'll add it to the to-do list. Actions will come from a variety of 'action' servers, but these IPs will be published and won't change.
@adly the SLA was to check on the monitored IP if it falls less than the threshold over a given time, so it helps the user file ticket for credits with his/her provider.
@peppr Got it. Though you could setup an action to automatically email the provider to request an SLA once downtime hits a threshold with the actions system too. Take all the work out of it. :-)
Im using serverstatus for uptime monitor its okay to look at but what would be more useful is to also add email alerts for up/down servers.I have to run a sperate script for that. I have those email alerts text my phone.
I don't care for too many options and features unless it's easy to setup and optional if you have a install script. I tried setting up nagios and similar services but didn't have enough time to finish setting it up and purged it.
Email/API/SMS/actions will all be possible responses to a down event. Simplicity is also key and additional features will be of course be optional. However, it seems you're talking about a hosted script, which this won't be.