NodeCheck (Beta) - Status Page, Server Monitor, DNS and Domain project

Franzkafka · January 25

looks great,I'm in.

Franzkafka · January 25

OK,so now the Affiliate program is not activated? "Invite people to NodeCheck with the following link and get 0% from their first payment to the platform." Magic number:0%;

And I found the UI in Mobile is not suitable when I create notification handler,you know what I mean here:

webmash · January 25

@Franzkafka Good spot, were just rolling out the Affiliate system - your just too quick for us 😀
Yes, the mobile version is a WIP currently, hope to have the mobile version working fully in a few weeks.

webmash · January 25

We have added more notification handlers for notifications so not only do we send an email, but you can also have a notification to: Telegram, Slack, Discord or a Webhook. Should have MS Teams, and SMS shortly.

TheDP · January 26

Name looks familiar

Well done 👍🏼

arirang · January 26

Seems not working well
Every 180 second but every 4-7 minute.

webmash · January 26

@arirang can you drop us a DM or email and we'll take a look for you. Email at bottom of our website

Otus9051 · January 26

Just signed up, will report in a day or two of usage.

UI looks great on mobile, props to that.

lorian · January 26

So first of all, thanks for this service.
But there are a few things I miss. So here is my wishlist:
1. You are using Contabo for your probes, who wouldn't be my first choice networkwise.
2. Your mailserver is leaking its IPv4. (DDOS-protection)
3. It would be nice being able to check for strings (like an "OK" on some /health) on the HTTP(S) check.
4. Please add Pushover. Even the big tech is using them.
5. Curently, I only see one request every 3 minutes. As there might also be network-problems, it would be nice to have at least 3 checks from different locations in one time frame.
6. And last: Please support IPv6.

Disclaimer: I am using Zabbix, so I probably won't be a paying customer in the future.

webmash · January 26

@lorian said:
So first of all, thanks for this service.

You're welcome - its still a WIP

But there are a few things I would change. So here is my wishlist:
1. You are using Contabo for your probes, who wouldn't be my first choice networkwise.

We don't just use Contabo and were testing lots of others, looking at the results and then putting them into production.

Your mailserver is leaking its IPv4. (DDOS-protection)

Yup, were moving - happening at the moment.

It would be nice being able to check for strings (like an "OK" on some /health) on the HTTP(S) check.

Please add Pushover. Even the big tech is using them.

Its in the works

Curently, I only see one request every 3 minutes. As there might also be network-problems, it would be nice to have at least 3 checks from different locations in one time frame.

Click on Edit on your monitor, then Advanced - you can choose 10 other locations

And last: Please support IPv6.

Being worked on currently.

Jab · January 26

What is the idea behind those labels? This makes no sense atm.

The status page URL Alias need to have checks on the plan - just slapping disabled in frontend is no go ;')

I've (most likely) broken monitor number 13 by trying different manual POST params, wanna investigate/fix it?

Feedback about status page:

There are dates (day-month) missing from labels on status page (they are still there in single server view) - this is last 7 days view of one of my servers on status page:
https://i.imgur.com/q6U7WoH.png
This is the same server with single view:
https://i.imgur.com/KYQohpd.png

Things are totally broken when there is no checks in database - seems like you don't handle that and don't add "empty/broken" entries on timeline.
This is monitor 13 status page view for last 7 days: https://i.imgur.com/DGPyzyw.png - this totally suggest the server is online and working... but those measurements have 21:00 in legend/right side of graph... and it's around 14:00 - those are 3 days old measurements getting pushed to the right.
This is same monitor 13 - single view: https://i.imgur.com/BR0SzJI.png - dates are here in labels.
Todays view: https://i.imgur.com/31YFGxx.png

webmash · January 26

@Jab said:
What is the idea behind those labels?

Just labels for us, they have been removed.

The status page URL Alias need to have checks on the plan - just slapping disabled in frontend is no go ;')

Its a WIP, lots going on in the background.

I've (most likely) broken monitor number 13 by trying different manual POST params, wanna investigate/fix it?

Yah looks like you managed to break monitor 13, we've removed it from your account, and looking at all the log and codes to prevent this from breaking again.

Jab · January 26

@webmash said: Yah looks like you managed to break monitor 13, we've removed it from your account,

Did you also removed the IPv6 one and Oracle that I had there for testing or they just somehow hidden/filtered/missed(?) because never turned online?

Otus9051 · January 27

Its inaccurate to some sense. I have 2 instances of Uptime Kuma running on Azure and AWS. While they don't report any outages, Nodecheck manages to always find one. It's weird. I wonder if it has something to do with the pings, because all of my nodes are atleast 150ms from the server (as indicated by the average response time).

Though I tried adding Google as a control, and as it seems it has 0.6ms average response time and hasn't yet had an "outage".

lorian · January 27

First of all: Thanks a lot for your response.

@webmash said:

@lorian said:
5. Curently, I only see one request every 3 minutes. As there might also be network-problems, it would be nice to have at least 3 checks from different locations in one time frame.

Click on Edit on your monitor, then Advanced - you can choose 10 other locations

Yes, of course. I did so, but (at least yesterday) I could only see one HTTP request in total every 3 minutes coming from different locations. So not every location was checking every 3 minutes, but only one round-robin one was.

One silly example: My hoster receives AMS and DUS via IXPs, but the rest only via transit. That's why I would add checking from this two locations aswell. Now, the DUS' IXP's DDOS protection feature is faulty and is blackholing my traffik, while the transit is working fine.
With every location performing a check every timeframe, it would trigger immediately/after 3 minutes, while in your current setup it could take half an hour for the DUS probe to perform the check and alarm.

Of course: 10 parallel checks might not be necessary, but 3-4 different locations performing the check every timeframe would be nice.

webmash · January 27

@Otus9051 said:
It's inaccurate to some sense. I have 2 instances of Uptime Kuma running on Azure and AWS. While they don't report any outages.

Thanks for the feedback, very much appreciated. I've looked at your account/monitors and see that AMS was having a wobble, so we've taken it out the mix for now, I've set your monitoring locations from the default London, let it run for the next few hours and then we'll check the log. Google is 0.6ms from London (possibly peering in place, or within the same datacenter

webmash · January 27

@lorian said:
First of all: Thanks a lot for your response.

@lorian said:
5. Curently, I only see one request every 3 minutes. As there might also be network-problems, it would be nice to have at least 3

Can you DM me your username, and we'll look at the logs as I can't locate you, but happy to have a look.

rsk · January 27

@webmash if you are looking for additional locations, we are happy to sponsor

Otus9051 · January 27

@webmash said:

@Otus9051 said:
It's inaccurate to some sense. I have 2 instances of Uptime Kuma running on Azure and AWS. While they don't report any outages.

Thanks for the feedback, very much appreciated. I've looked at your account/monitors and see that AMS was having a wobble, so we've taken it out the mix for now, I've set your monitoring locations from the default London, let it run for the next few hours and then we'll check the log. Google is 0.6ms from London (possibly peering in place, or within the same datacenter

Yeah I was wondering how was my HA node (August) getting downtime
Thank you!

webmash · January 27

@Otus9051 said:
Yeah I was wondering how was my HA node (August) getting downtime
Thank you!

Welcome, we're fixing a few things this weekend London Monitoring is perfect now so that's been put into production and performing well. We'll have the others ready this weekend. Notifications, email also now dispatched faster via MB/MC & SES. Contabo dropped (tests were terrible, and network was patchy) IPV6 testing also taking place this weekend.

Big shout out to the LES community for all the feedback and support, its greatly apprecaited.

Otus9051 · January 27

@webmash said:

@Otus9051 said:
Yeah I was wondering how was my HA node (August) getting downtime
Thank you!

Welcome, we're fixing a few things this weekend London Monitoring is perfect now so that's been put into production and performing well. We'll have the others ready this weekend. Notifications, email also now dispatched faster via MB/MC & SES. Contabo dropped (tests were terrible, and network was patchy) IPV6 testing also taking place this weekend.

Big shout out to the LES community for all the feedback and support, its greatly apprecaited.

Bit of a design criticism,
On the sidebar, the color of it is very close to that of the "node" in the logo, impairing visibility.

lorian · January 29

@webmash said:

@lorian said:
First of all: Thanks a lot for your response.

@lorian said:
5. Curently, I only see one request every 3 minutes. As there might also be network-problems, it would be nice to have at least 3

Can you DM me your username, and we'll look at the logs as I can't locate you, but happy to have a look.

Sorry for the late reply. You just got some logs.

ialexpw · January 29

@webmash said: Welcome, we're fixing a few things this weekend London Monitoring is perfect now so that's been put into production and performing well

How many checks does a location do before reporting a site as down? I had 6 downtimes reported in a row on the 26th and I had to disable the monitor due to continuous emails (down/up/down etc). Does it check via another location at all?

It's a nice looking site though, I'll come back and try again at some point.

Joseph · January 29

Will this become a paid service?

webmash · January 29

@Joseph said:
Will this become a paid service?

Not for a while, we're still in beta and have a lot to roll-out still. Eventually we will offer paid accounts and free accounts.

webmash · January 29

@ialexpw said:

@webmash said: Welcome, we're fixing a few things this weekend London Monitoring is perfect now so that's been put into production and performing well

Could you DM us your username on NC so we can look at the logs please

webmash · January 30

Good morning from NCHQ, couple of system updates for you this morning regarding monitoring locations. NodeCheck uses a distributed monitoring system to minimise false-positives.

All primary checks are now made from London, United Kingdom as default. However, once a downtime is detected, secondary requests to verify this downtime is sent from remote nodes that are located in different countries/continents. You can of course change and add locations when setting up a monitor click Advanced and more options are now visible.

Server monitoring has been updated and now includes, CPU/RAM/DISK/NETWORK and Other stats - https://nodecheck.net/server-monitors If you have it installed currently use the uninstall code, and reinstall the code to get the updated logs. (working on an auto updater)

This week were working on custom domains for status pages and more status page updates and layouts.

lorian · January 31

@webmash said:
All primary checks are now made from London, United Kingdom as default.

Then you still have the problem, that partially network outages won't trigger an alarm, when the link to London is fine.
You may want to check from at least 2 sites (London & US?) or even better from 3+ sites (EU, US, Asia) with an alarm, when n-1 sites are detecting an outage.

webmash · February 1

This week were working on custom domains for status pages and more status page updates and layouts.

webmash · February 1

📣 Last call for beta accounts before registrations are closed. Grab an account https://nodecheck.net

NodeCheck (Beta) - Status Page, Server Monitor, DNS and Domain project

Comments