New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
The v2 on the picture has an Allwinner H313 with 2GB RAM. The firmware is Yocto based, we will just update the existing one https://github.com/jsdelivr/globalping-hwprobe
In theory it should be a pretty good fit, but we will see after testing
eyeball and datacenter tags are live https://www.jsdelivr.com/blog/new-globalping-feature-eyeball-and-data-center-network-tags/
There are a few false-positives with Misaka, Oracle and a few other networks but we're working on fixing it.
And we open sourced our dashboard https://github.com/jsdelivr/dashboard
Currently focusing strictly on the backend functionality and later will start the design polish process.
Anyway, I hope it works out for you,Good luck.
Some updates:
Everyone running a hardware probe should consider flashing the SD card with the new firmware By doing so you will also check and ensure that the SD card is still alive. (New ones will be more reliable.)
[@1 | first, @2 ... @-2, @-1 | last | previous]Web tool improvements:
We made many changes and improvements to the web UI, you can check a PR preview over here https://www-jsdelivr-com-pr-624.onrender.com/globalping
faaaaaaaaaaaak, I knew I forgot about something and I've never bought new SDcard.
Flashed my probe, tanks for the reminder.
The UI changes were merged and are live https://www.jsdelivr.com/globalping
And I wanted to share this security review of the Globalping probe which I think is a cool read
https://dspace.cvut.cz/bitstream/handle/10467/115701/F8-BP-2024-Tyukos-Viktor-thesis.pdf
I'm told that when Globalping is requested to MTR toward two target IPs, the latency toward the second target is always higher.
What's wrong?
https://lowendspirit.com/discussion/comment/201307/#Comment_201307
I was looking for EXACTLY this just the other day. Starred your project and installing a node now.
We're looking into it, probably a frontend bug. Thanks for mentioning me here, for some reason I didn't get an email notification from the spirit forum.
The bug was fixed, thank you for reporting this! https://github.com/jsdelivr/www.jsdelivr.com/commits/master/?since=2024-12-27&until=2024-12-27
We have lots of features planned for 2025, but if there anything else we're missing or something doesn't seem to work properly please let me know!
We've been working hard on Globalping and here are the most interesting updates:
Your own probes are highlighted in the CLI.
Public profiles are live (opt-in in the settings) https://globalping.io/users/jimaek
If you enable them then others can also target your probes by username
Auto-adoption flow is now live. You can auto-adopt your probes by simply setting an ENV VAR with your unique token
TLS data on the website and CLI
n8n integration https://github.com/jsdelivr/n8n-nodes-globalping
And more features are coming! You can always join our GitHub issues to provide feedback and ideas and monitor the releases of new features.
Wait, so average Joe, what can I do with this tool? Is it only ping or can I get some info about IP in reliable way?
@jimaek a few things I would love seeing after adding like 300 probes lol:
now i can ping all from my probe at the same time
This is massive improvement tho!
Its not a geo IP tool. You can use to benchmark providers, check for uptime, do latency tests. You could in theory even try to detect the location of an IP using latency data and some math.
But this is just the beginning, later we will start using this infrastructure to provide data like rankings of CDN, DNS and Cloud providers.
Makes sense, we will add it
It's coming https://github.com/jsdelivr/globalping-dash/issues/66
Can you PM me the IPs?
Will be coming but a bit later.
And thanks for hosting so many probes!!
I'm running 2 hardware probes and about 10 docker containers. In which countries you don't have probes/insufficient probes?
Latin America and Africa are always a problem. But the more probes the better, even in established places like Europe. We need more ISPs and ASNs.
You can explore our map in detail over here https://globalping.io/network
hey @jimaek i wonder if there's a plan for a personalize globalping probes stats? or it can be public, like how much request does this probe performs in the past day, months etc. and what types of request something like that.
Satisfaction stats i'd say
Yep, something like this
But we first need to launch our timeseries DB backend to handle the data. Coming this year.
I can run a hardware probe in Africa but I'm gonna tell you it's 95% gonna be stolen by our customs or our post office.
haha you can always use a Docker container on an available computer
Big fan, gonna make my auto-deployment even more auto.
Just added all my VPS, auto adoption is great
feature request: binary additionally to container installation. no need to install podman / docker on some low spec idler
(see hetrix, beszel, etc.)
Awesome! Thanks!
Unfortunately it's not possible. The container comes with many tools that we need like the correct versions of binaries for traceroute, ping... It also allows us to easily auto-update the code inside the container.
It's critical for the system for all probes to always be on the latest version.
This is also a major blocker for us to simplify installation on Windows machines
But I'm hoping I'll be able to crack the hardware probe topic and create something with a screen that doesn't cost too much. That would provide more value and would be easier to promote to less technical users.
Can someone help me diagnose a network issue? Im trying to figure out how to troubleshoot this issue.. my ISP ASN can't access/communicate with my server not sure how to troubleshoot this, but when i use a VPN i am able to get a response from the server.
Running a couple of probes on "unique" ASN's.
Great service. But I wish you were still able to perform tests without having to login, and with a decent amount of credits.
What do you mean? You can run tests without login in all of our tools with a 250 tests per hour limit. We had to set limits low enough to avoid abuse and high enough to still be useful.
Interesting. Last time I checked it was 0 credits. Maybe there was some outage.
I find 250 a bit too low, its fine, but I wish it was something closer to 1k.
I rarely login to my account, even though I have a lot of credits stacked. Prefer to use it anonymously.
I wish there was some auth option other than Github, and If added, an option to migrate from Github auth.
Got it, we will look into adding more auth providers. But FYI if you're not authed the limit is applied per IP address. So if your home IP hits the limit you could install our CLI on a server and continue there with a new limit, without auth.