uptimerobot | Client library for UptimeRobot v2 API | REST library

For uptime robot to work, you need a publicly facing webpage that it can request every 5 minutes.

Two options:

1. Your discord bot needs to run in the cloud and run a http server at the same time

2. Have your web server accept another request like example.com/keepalive to tell the server that the bot is still running.

Uptime robot would then query example.com/alive and you have to return 200 OK if the bot is online and 404 if offline.

NOTE: If you already have something like the 2nd approach running, you may as well just save the uptime info on your public server. I also recommend passing some authentication in headers or as a query (password, for example) if you use this to avoid people abusing it.

A HEAD request should respond with only headers, not content. So mixing with a GET shouldn't be done.

The easiest option is to create an extra method in the controller to handle that HEAD request:

Unfortunately, there is no way to add more than 10 custom domains to a SSL in AWS. Which is a prerequisite to having a custom domain to your cloudfront.

Hence, a workaround this could be as below. 1. Create a S3 single bucket which hosts your code 2. Create Multiple Could front distributions connected to single S3 Bucket 3. Then, add custom domains to these cloud front.

You will also need to think about CORS settings in your API of the app to allow requests from these custom domains.

In your terminal, require the guzzle package as you will use it as an HTTP client: composer require guzzlehttp/guzzle

Then you can make a class for your UptimeRobotAPI at app/Services/UptimeRobotAPI.php:

I have frequently seen the one-minute failures on low-traffic instances, such as testing environments. Every time I've investigated, the 4XX errors were from a port scanner or some other malicious requests. Since traffic on non-prod instances is low, it doesn't take much to trigger "85.7 % of the requests" - that could be as few as six out of seven requests, for example.

You might see the 4XX errors in your ELB logs if they aren't showing up in your application logs. ELB logging is disabled by default, but you can turn it on and log to S3.

The simplest approach would be to restrict access to your application by whitelisting IPs in your security group. If your application needs to be public-facing, however, then you have a few options for addressing the problem:

1. If the requests are coming from a single IP address, you can block it using an ACL in your VPC.
2. If the requests are coming from multiple IP addresses, you might be able to block them if there is any consistent pattern, such as what URI they're attempting to access, the associated User Agent, or the like. You'd need to enable WAF, however.
3. Just ignore the warnings - they're most likely harmless, and once you have more traffic, they'll blend in with the rest of the noise.

get() returns a collection of all of the rows, since you're getting just one column and not specifying a where clause. If you only want one value, you can use value to get a single value (previous versions used pluck)

I can suggest slightly different solution:

1. Extract your curl code in a separate console command and run this command each minute (for example, as a cron job).

2. The result of the command save to database/file/memory.

3. In your showMonthlyReport() refer to existing result.

Benefits:

• In this way you would not have to wait for your curl result on each showMonthlyReport(). All code will run asynchronously
• All errors processing will be in one place
• Command is testable

So, I managed to solve it. I asked for help on the HangFire forums and finally did it. Here's the link to the post: https://discuss.hangfire.io/t/keep-alive-on-shared-server/3723

TL;DR: Basically, I'd tried ping and HTTP requests with uptimerobot and they didn't work for me. What did was their Keyword requests. And now I have a windows scheduler simulated service. :)

Visit the HangFire forums for details.