0

I am working on a website https://devsdata.com/. It is hosted on the Google Cloud Platform. We are using "Better Uptime" to monitor its health and every couple of days we receive an incident alert stating that the website is down. After being down for 4-5 minutes it then goes live back again without any work required. What could be the reason for this? Should I look into DNS? The server has enough free space to operate normally and CPU/RAM is more than enough for our typical load. It is a WordPress website created from the Openlitespeed Google Cloud package.

5
  • Does each outage occur at the same time of day? – ceejayoz Sep 30 '20 at 13:12
  • @ceejayoz no, it doesn't. It occurred on 18, 25, 26, and 29 of September at 8:44 PM, 5:57 PM, 4:55 AM, and 9:33 PM respectively. – Nikita Grygoriev Sep 30 '20 at 13:19
  • You'll want to consult your access and error logs around those times, then. Perhaps it's a badly configured crawler/scraper accessing the site or something. – ceejayoz Sep 30 '20 at 13:26
  • @NikitaGrygoriev you mentioned that your website is hosted in GCP, Could you let us know what Machine type your are using in your VM Instance? did you use a deployment to create your Wordpress site? Also, have you checked the logs for your instance on the Logs Viewer page please share with us as much information as possible – Jose Luis Delgadillo Sep 30 '20 at 13:28
  • @JoseLuisDelgadillo machine type is n1-standard-4 (4 vCPUs, 15 GB memory). Yes, I have used a deployment to create it (openlitespeed-wordpress deployment). The problem with logs is that the machine does not create any at the time the downtime occurs. The only log from the latest downtime is 12 hours off and as for the downtime on September 26th, there are no logs for that date at all. – Nikita Grygoriev Sep 30 '20 at 13:35
1

If you want to discard the DNS issues, start monitoring the IP instead of the domain.

If you are monitoring the domain instead of the IP you are not monitoring directly the Instance, and there could be a lot of factors involved.

Also, it seems that the only way to monitor that has Better Uptime is through ping, and maybe port scanning but I'm not sure, they are not clear on how it works.

Additionally, I consider that you could take a look at this page: Creating an uptime check it explains how to create and configure uptime checks. You could create an uptime check to corroborate if your site is down. I think it is a good idea to monitor the instance inside in order to compare the information you are receiving from Better Uptime with the information from your instance.

5
  • So I have created an uptime check as you recommended and have received an alert: Failure of uptime check_id devsdata-website-uptime-check Violates when: Any monitoring.googleapis.com/uptime_check/check_passed stream is above a threshold of 1 for greater than 1 minute I did not manage to visit my website while this alert was taking place, but about 1 minute after receiving this alert the website was working perfectly. – Nikita Grygoriev Oct 1 '20 at 14:42
  • Could you connect the events from your tool with your web server logs? It is not clear if your website is down, or if another factor is getting involved for example it could be a problem in the DB, or a plugin from Wordpress, that makes a status different to OK, Additionally, you could check your Uptime-check policy, it might not be adjusted correctly, you can take a look at this example: cloud.google.com/monitoring/alerts/policies-in-json#json-uptime. – Jose Luis Delgadillo Oct 1 '20 at 20:00
  • These are the configurations of the uptime check: Check Type HTTPS SSL Validation Disabled Path / Port 443 Check Every 60s Timeout 10s Regions Global And here are the logs for when the uptime check failed: pastebin.pl/view/680bdefd Could you take a look at it? – Nikita Grygoriev Oct 2 '20 at 8:09
  • Did you match the information from your Uptime Check with the logs in your instance or with the information provided by Better Uptime?, I think it could be a good idea to have all the information in order to determine what could be the issue. Another option is to View the serial port one output to get more information about what is happening in your instance at the moment of the issue. – Jose Luis Delgadillo Oct 5 '20 at 22:29
  • Also, if you have a Free trial account you can have chat support through Console Support Center, also, you can visit the following link for more information. – Jose Luis Delgadillo Oct 5 '20 at 22:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.