New to the community? Start here

Option to change the uptime monitoring time period longer

I have switched over to using WPMU to monitor uptime. I was using Jetpack.
But I’m getting a lot of false positives. Is there anyway to change the timeout settings so it waits longer or makes more checks. I was just wondering if I can control the settings someplace. I really don’t want to turn it off.

Fri, 24 Apr 2020 9:47:11
  • Adam
    • Support Gorilla

    Hello heartlandhosting

    I hope you’re well today and thank you for your question!

    It’s not possible to change timeout and/or frequency of checks, I’m afraid, but the timeout is already long enough (30 seconds) to assume that if site doesn’t respond within that threshold, it’s either down or slow enough to be pretty much “unusable”.

    That is, assuming that, those alerts are not “false positives” which might also be happening in some cases.

    To start with, what you could do to get less “false positive” reports, if you’re getting them over e-mail, would be to change “threshold” on notifications. You can do it on site, on “Hummingbird Pro -> Uptime -> Notifications” page in “Threshold” section.

    The time that you’re setting there means that only incidents longer than the set time will be reported to you over e-mail. So, for example if you set it to 30 minutes and the “down” incident will last 5 minutes – you won’t get any notification. If it will last 31 minutes – you’ll get notified.

    That’s a “workaround” to “silence” some of those false alarms but could be a good start.

    Second thing to do would be to make sure that our Uptime check is not blocked by site and/or host. It’s not unusual that some hosts “limit request rate” which might result in a lot of such false incidents. In a nutshell:

    – uptime is pinging site every 2 minutes
    – host allows that but detects continuous pings over longer period of time
    – and at some point starts to either completely reject them or just “silently ignores them” (which would be visible on our end as a plain “timeout”:wink:

    I would then suggest to make sure that at least following IPs are white-listed an every possible place – any security plugin on site, any external security suite if there is any (such as e.g. sucuri, cloudflare or similar) and in any firewall/request limiting configuration on server (so you might want to contact your host about that): 

    34.196.51.17  
52.57.5.20

    If none of that helps, it’d be recommended to actually check the site to see why it doesn’t respond in time and how to possibly fix it so in such case, let me know here, please, and we’ll investigate it further.

    Best regards,
    Adam

  • heartlandhosting
    • Design Lord, Child of Thor

    Thank you Adam.

    This is all very helpful. I will look into all these. I believe my biggest problem is with cPanel and the cPanel app, Jetbackup. I currently have this deployed across 3 VPS. Jetbackup creates the files locally and then copies to Dropbox before deleting. The “false positives occur only during the backup process. At this time I’m transitioning away from cPanel to Gridpane which is exponentially faster and more secure. So I’m hoping that everything with WPMUDEV is Nginx compatible, as I love and use your services for all of my hundreds of clients.

    Sincerely

  • Adam
    • Support Gorilla

    Hi heartlandhosting

    Thank you for response!

    The “false positives” occur only during the backup process.

    This is actually a significant finding. I’d say that it means that it’s a “resources issue”. Backup and related processes are usually, in general, quite “resource hungry” and very likely what happens is that this process takes so much resources that the server indeed slows down to the degree when it just can’t respond on time to Uptime requests.

    It doesn’t necessarily mean that visitors are not able to load the site at that time because a lot also depends on server configuration and how it handles requests. It’s quite probable that a real visitor will only experience some slow down during that period while Uptime request will stay “queued” for longer – long enough for Uptime service to consider that it didn’t get response. In fact it’s even possible that server does actually respond to such request but just later than within expected 30 seconds – such response wouldn’t be logged on our end.

    So I’m hoping that everything with WPMUDEV is Nginx compatible,

    It should all be fine. Most plugins don’t really “differentiate” whether it’s Apache or Nginx or other server but e.g. in Hummingbird we do actually have some nginx-specific configuration options so if your site’s powered by nginx you can simply select nginx wherever “server type” has to be selected/is relevant.

    In fact our own WPMU DEV Hosting is fully nginx-based so I believe that’s the best proof that our plugins/services can be used with nginx :slight_smile:

    Best regards,
    Adam