Uptime Monitoring is way too sensitive

Hey there,

Just wanted to tell you that you seem to have a problem with false positive for your uptime monitoring feature, at least on my end.

I keep receiving emails telling me a website is down (all websites, not a specific one), so I check and it's not down (but I don't get the "website is up again" mail in between, I only get this one much later).

I've been checking this issue many times on many different websites before reporting it.

It's not a big deal, but it's a shame, because with so many false alarm the uptime monitoring doesn't mean anything anymore, and become useless.

Do you think there is a technical problem somewhere, or is your monitoring just way too sensitive ?

Isn't there a way to at least adjust the sensitivity, so for exemple I can be warned only after a 15 min or 60 min downtime ?

Cause sometimes I receive like 20 emails telling me my websites are down, and they aren't when I check...

Any thoughts ?

Thanks a lot,

Dushan

  • James Morris

    Hello DJExp

    I hope you are well today.

    The way Uptime currently works is it "pings" your homepage (connects would be a bit more appropriate) at predefined intervals (about every 2 minutes). If the connection takes longer than 30 seconds for the site to respond, this returns a failure in Uptime. Other sources of failures in Uptime are any non-200 response. ex: 30x redirect, 40x response, 50x response. Most commonly we see members running Maintenance Mode plugins which set the header response to 503 which will cause a false positive.

    If none of the above causes for false positives apply to your site, then it is likely that our system is registering that it's taking longer than 30 seconds to get a response from your server. This could be due to DNS issues or a high TTFB value.

    The best way to determine what could be causing this would be to use tools like Pingdom Tools and Dotcom Tools to determine if there are any bottlenecks that need to be addressed. It's entirely possible that the sites will load fast for you, but respond slower elsewhere in the world.

    At present, there's no functionality to adjust the timeout or poling interval for Uptime. If you would like, I can add this as a feature request for our Dev Team to consider for future versions of the service.

    Let us know if you have any further questions. We'll be happy to help! :slight_smile:

    Best regards,

    James Morris

  • DJExp

    Mohamed : Thanks ! Where should I whitelist these IPs ? On what Firewall ?

    James Farmer : Thanks for the detailed answer.

    Yes, I get the "false positive" for websites that are in maintenance mode, but that's normal. What is weird is that I get a false positive many times a week, and when I check (sometimes IMMEDIATELY because I happen to be facing my inbox when I get the message), the website is reunning fine and responding quickly.

    So what can I do to improve the diagnosis ? Should I use the 2 tools you mentionned when WPMUDev Monitoring is showing my website as offline ?

    Last but not least, yes, adding the timeout and pinging interval for the website would be an awesome addition ! You don't need to ask my permission to add this as a feature request : if you think it's a good idea, go ahead. I'm glad that I can somehow contribute to improve WPMUDev tools, because those tools helped me so much over the time ! :wink:

  • DJExp

    Hi James,

    The problem with the 2 tools short term solution is that it doesn't solve my main problem : being overwhelmed by false positive.

    I get so many wrong uptime emails that I just delete them without checking now (because 90% of the time it's a false positive), so I'm pretty sure that the one time I'll get a real problem I won't pay attention to it.

    Can't you think of any solution or workaround that could prevent it, other than simply disabling the uptime monitoring completely ?

  • James Morris

    Hello DJExp

    I'm afraid that, at this time, there's no way to adjust the timeout threshold for Uptime Monitoring or to change the notification frequency. Work is being done to improve this, but is not ready for production yet. So, until these changes get pushed to production, I'm afraid that, if you're getting too many emails and you know for sure your site is up, the only solution would be to either re-evaluate your hosting and upgrade to a faster host or disable Uptime for now.

    Best regards,

    James Morris

  • DJExp

    Hi James,

    Just to let you guys know that after receiving dozens and dozens of message telling me my websites were down and up and down and up and down, I deactivate Uptime Monitoring on my whole network, because it's just ridiculous, it wastes me time deleting emails and it completely defeat the purpose : assuming one of my website was REALLY down (which it wasn't), I wouldn't even pay attention to it because I learned that 99,9% of alerts are false positive.

    Also, my host is very powerful, it's one of the best in Europe (Infomaniak), so I doubt it's the problem.

    Please do let me know if you fix this service by implementing a way to adjust the timeout threshold, or simply by avoiding fasle positive, and I would be glad to test it again, but for now it's just a big waste of my time. Thanks

Thank NAME, for their help.

Let NAME know exactly why they deserved these points.

Gift a custom amount of points.