Can you disclose how your assemble your GeoIP information? It doesn’t seem to match Maxmind’s for the couple of IPs I’ve tested.
One small piece of feedback: the CIDR pages don’t make much sense to me. For example, https://worldip.io/range/1.0.0.0/8 shows as Cloudflare owns the /8, when they actually own the /24, /8 is not allocated. I understand those pages are generated, so please consider something similar to bgp.tools’s prefix not found page, which shows me all allocated ranges, including sub-allocations: https://bgp.tools/prefix/1.0.0.0/8
I've recently started aggregating my own location data to run alongside MaxMind. If you don't mind sharing a few of the IPs that were off, I'd love to take a look and see if there's a dataset bug.
Good call on the range pages, too—I’ll look into handling unallocated blocks more like bgp.tools. Appreciate you bringing it to my attention!
In a similar tone, I made Caddy module that temporarily blocks an IP address once it tries to access specific URLs: https://github.com/kassner/caddy-trapdoor It’s great to catch those bursts of requests, nowadays I see a single IP probing hundreds of paths in a couple of seconds. This catches majority of them for me.
It depends how much EU do you need. ZeroSSL is through HID Global, part of Assa Abloy, a Swedish public company. HID, however, is American according to Wikipedia. Caddy uses ZeroSSL by default.
You wouldn’t be charged thousands because the service is pre-paid. Your websites will be suspended once the balance runs out, but you can re-establish them once you add more balance.
AFAIK is the only provider in which you can have functional billing limits and not just alerts that still depend on you reacting on them in time.
Where in the terms of service or in the admin UI do you see this information? You can pre-pay, yes, and the CDN service (and ONLY the CDN service) has a setting to stop the billing. But the other services do not seem to have a limit on how much you can get billed.
But yes, other services will run until your account is dry, then it gets suspended. You can't set a spend limit on your scripts without it taking down the whole account. You can't do cost optimization, but at least you are protected from going bankrupt.
It’s not like they’re going to put a debt collector after the $0.73 that you’d owe, so I’m not sure what was your point. If you want the services to continue you gotta have some balance. If they allow your account to run below balance I’d say that’s on them.
It's not an accurate correction. It's prepaid with an ability to incur a small amount of negative balance, partly for protection against downtime or data loss and partly so the systems can have some time to shut off (it would be hard to suspend service immediately across all services). However it is purely prepaid in that they don't provide a postpaid option.
That's a very uncharitable way of dismissing my claim that the parent claim of the service stopping automatically when you go over your credits is not true.
That's one thing. The second thing, I'm happy for you that you have more information about their pricing system than their pricing page and terms of service page state.
It's inaccurate to say that it isn't purely prepaid, because that could be taken to mean that some services are available to be postpaid, with a due date. Here you can see that a negative balance isn't the same sort of thing as having a postpaid bill due: https://support.bunny.net/hc/en-us/articles/8726637076124-My...
It's not not purely prepaid. It is purely prepaid through I might favor another way of saying it that specifically says that they don't offer any services on a postpaid basis. But even if you define purely prepaid differently from me, it doesn't work so well as a correction because the comment being replied to didn't use the term "purely prepaid" or something highly similar.
Ansible will be terrible slow to update things if you're geographically far from the servers. Updating around 100 servers from another continent already takes longer than a 5 minute TTL. Parallel connections only work up to a point before a playbook run always have at least one failure and requires a re-run.
Ansible also does not have locks or parallel users coordination, so you’ll need a single user/VM/GHA workflow running the playbook or at some point concurrent users will start overriding each other.
I wonder if at some point we’ll be start taking about non-smartphone sovereign payments. The main reason I still use card is to be able to use it without a phone, and the technology of debit cards (around Europe at least) is quite OK. Maybe Europe should have a parallel payment track that is just a new card brand.
It is pretty much unacceptable to have a domain bouncing emails, so I’d be out of the provider before the MX TTL even expires.
For outgoing emails, reputation is a huge issue, but at the same time it’s also fairly trivial to set up a (different) 3rd-party (gmail, outlook, sendgrid, whatever) with previous reputation so you can get back communicating.
Can you disclose how your assemble your GeoIP information? It doesn’t seem to match Maxmind’s for the couple of IPs I’ve tested.
One small piece of feedback: the CIDR pages don’t make much sense to me. For example, https://worldip.io/range/1.0.0.0/8 shows as Cloudflare owns the /8, when they actually own the /24, /8 is not allocated. I understand those pages are generated, so please consider something similar to bgp.tools’s prefix not found page, which shows me all allocated ranges, including sub-allocations: https://bgp.tools/prefix/1.0.0.0/8