GeoIP with ipinfo stopped working ?

[reincoder]

Hi,

I am Abdullah, the DevRel of IPinfo. I will try my best to help you here.

Our file format will never change to anything else without a significant amount of communication. Please understand that a file format change without alerting our user base is a catastrophic change that we will never make. So, I suspect this is some error message being interpreted as zip file or something.

I am not clear about the implementation of how the database is downloaded, but we do have a checksums API endpoint that you should use to verify the download.

Reference: https://ipinfo.io/developers/database-download

The download process requires you to go through a redirect path because the data is stored in a cloud storage bucket.

Reference: https://ipinfo.io/developers/ipinfo-lite-database (See the code section)

The API provides unlimited usage, but the data downloads are subject to rate limits. It permits 10 downloads per unique IP address multiplied by unique access token. This means that to reach the rate limit, you probably downloaded it 10 times using the token you are using from the same IP address.

Reference: https://community.ipinfo.io/t/announcement-we-are-adding-rate-limits-to-data-downloads/358

You have shared your API access token: `f2cbc8898bc30a` which according to our database is a not an active or assigned token.

---

Please let me know if this problem persists. We will be happy to take a look. Our community forum is available here: https://community.ipinfo.io/

— Abdullah | DevRel, IPinfo

[reincoder]

the only thing I can assume is that ipinfo removed the "Content-Disposition" header ( it's hosted in cloudflare it doesn't make sense ).

Let us investigate this issue. I have escalated this to engineering.

— Abdullah | DevRel, IPinfo

[reincoder]

Hi,

We were slowly migrating to a different cloud storage service over a period of time. We have rolled back the system migration entirely.

We have been doing incremental migration throughout the process, and thanks to the post and the only message we received on the support portal, we discovered reliance on the URI header metadata. This was not a hard transition, which has been rolled back, and further investigation will be conducted in the next few days.

During this transition, thanks to this post, we found out that the Content-Disposition header was no longer included in the final download response. This header was optional, but OpnSense relies on it to detect the filename or file type. We have been doing the slow rollout for several weeks, and this was the first issue we received.

The file itself has not changed; it's still a .csv.gz gzip file, but because the header is missing, some scripts may incorrectly treat it as a ZIP instead of a GZIP file. We have rolled back our entire deployment to our previous storage service.

We will investigate closely in the coming days. Thank you for understanding.

— Abdullah | DevRel, IPinfo

[SteffenDE]

Great to hear, thanks for the support.