All umami instances have been infected with a persisting crypto miner. Umami was affected by the next.js CVE but quietly released a fix, so most of their users missed it
You must log in or register to comment.
Link? Did you discover this yourself? There is no actual info here.
Thank you!
All recently open issues are about this. I was a victim, but I’m not the first and people on reddit have done better investigations than I have. Look for the name of the process at the top
Thanks.
For severe incidents like this, please post the most appropriate link, in this case https://github.com/umami-software/umami/issues/3852
Admins in self hosted usually don’t have that much experience with real, active compromise and may panic, let’s help them as much as possible.
I will add that Umami itself is not compromised, but vulnerable. That is a somewhat misleading title.
What was the vector? Did you have umami exposed publicly?
I don’t know about “all umami instances being infected” but they were certainly all vulnerable.
All umami instances have been infected with a persisting crypto miner.
Source for that claim? Because vulnerable does not mean infected.
Also, I’m kinda glad my instance has been offline for a while now because of database trouble. That was lucky.
Look inside
React2Shell
Just another day on the job
Wow I’m glad I happened to see this here. Thank you for the post. I was just thinking about putting all my services behind a VPN too, I think I’m going to go ahead and put that at the top of the list…
I don’t think a vpn would help here
Removed by mod
Yeah but Umami is an analytics engine powered by client side tracking. If it was behind a VPN it would be useless.
Yes I re-read the cve, I thought it was an issue with an npm package with a cryptominer
Unless it was the software package itself that was compromised.
It was not
This could explain why my 4C/8T VPS started hitting 100% CPU usage shortly after boot with like next to nothing else running on it.
Yup, umami was the culprit in my case. Quick update and it’s all running smooth again.
I see it’s running Ansible. That’s an obvious risk.
