All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
BunnyCDN has been silently losing our production files for 15 months, do not trust them for storage
On Jan 13th, 2025, we noticed strange 404 errors in our backend logs, originating from Bunny Storage. We investigated and found that files which were uploaded successfully to Bunny via their API simply vanished, with no deletion from our side, and no recorded write operation of any kind in Bunny's own logs.
Bunny's own support confirmed it the next day (January 14th, 2025) saying: files were found in the replication region but not in the main region.
Timeline
Jan 13, 2025 - Ticket opened after dozens of missing files.
Jan 13 - Support asks if the files were "deleted then re-uploaded." We explain they were never deleted by us and there are no logs showing deletions on Bunny's end either.
Jan 14 - Support escalated to their Storage team. They confirm the files were found in some replication regions but not in the main region, which is odd. Engineering will "try to recover files."
Jan 15 - More 404s.
Jan 17 - 200+ instances in 7 days. Four different agents have touched the ticket. None have a resolution.
Jan 19 - We ask to speak with someone directly. Refused. All communication must go through the ticket system.
Feb 2-6 - Still not fixed. Every response: "forwarded to the team, no update yet."
Apr 8, 2025 - Months of silence. We bump the ticket with new missing files. Storage team is "still looking into the longstanding issue."
Apr 24 - More files gone. "Escalated with the development team, no update at this time."
Apr 29 - Another lost file. "Chasing with the team."
Mar 24, 2026 - Nearly a year of silence. Bunny reaches out unprompted: "Our Storage team have not come to a conclusion. There were recent deployment changes to improve resilience." They ask if we have more recent examples.
Mar 26, 2026 - We report a file uploaded that morning is already gone. "I've updated our Storage team and we'll follow up."
Apr 9, 2026 - After bumping the thread again, support confirms the original bug still exists.
The loss rate isn't enormous in percentage terms, but it's consistent and ongoing. Yes, we should've migrated away sooner, we never had the capacity to do so and hoped Bunny would just get their shit together.
Bunny acknowledged this issue over a year ago. The files are recorded as sent to storage, are briefly available, and vanish into thin air hours later, with no recorded deletion or any other write operation on these files.
The infuriating part is there's nobody to speak with, we just get the same answers from Support with zero escalation options. Do not use Bunny Storage for anything you actually care about.
Happy to answer questions.
Comments
Please post this to hackernews at ycombinator. There folks considering bunny as alternative to cf (laughable).
@BunnySpeed
Have you got a backup or you used BunnyCDN as backup? Are you aware that you can store the same data in different BunnyCDN destinations?
Cost optimisation to the next level 🚀🚀
It's scary for me . I have signed up a few days ago and have planned to move a project to them. It is a good feature to get your objects stored at multiple locations, but there is no point when original files are being deleted on their own.
15 months is too long
At the same time, there isn't a lot of debugging information to base your investigation on... "Just disappearing", while worrying, isn't enough debugging info.
That's true, but it sounds like they've been told which files have disappeared, and one would hope that they have some level of logging on their systems in order to audit and correctly bill the customer.
Ou, we would be shocked to actually know how billing works at backend… stich here and there
Yeah silent data loss with no logs is nightmare fuel that’s the one thing a storage provider absolutely can’t mess up.
If there's 1% of chance files disappearing, that's enough reason to stop using the product. Unless it's more cost effective to log the missing file and re-upload from backup.
It would be helpful if @eran1243 shared what kind of files is disappears. Is it tiny files (below 1kb) or big (roughly a GB) or huge (more than a TB). Is it replication failure on other regions or inaccessible from any region
Files are images and videos - in the 1-10MiB range.
High priority? Clearly not high enough if it has been going on for that long.
From Bunny on Discord:
I kinda doubt the "isolated" part
@eran1243 Can you please post the percentage loss so it can be compared against other providers?
(I agree not knowing which file is lost is unforgivable though, since both S3 and B2 will display that (though not alert you unless you have a script configured))
If their priorities take 1 year... what's a non-priority, a decade?
I gave up on Bunny for very similar reason in 2023.
First initial ticket about 404's - ignored for 4 months, silently (auto?) closed.
Re-opened as different ticket later, there was some traction from low level support agent that kinda changed my CDN settings and asked if this is better and narrowed it down to "perma cache". Yeah, I told them it's perma cache related in 2022. Then ~6 months of "We have no update, forwarded your findings to XXX team" or once I saw "Team can't replicate" while I can replicate every time. Told them to close that ticket, moved to CloudFront.
In nutshell - if you had perma cache enabled bunny would just throw 404 sometimes out of blue and even not try to connect to origin server, at all. If you refresh the same URL 3 seconds later it would work. Kinda curious if OP bug is continuation / the same bug :-D
I have moved everything DNS related to bunny so I can have a cool custom nameserver but still kept everything else like CDN in cloudflare.
As much as I like Bunny and have been following their rise over the years. I found their support organisation to be subpar. Which meant we no longer went ahead with our implementation plan.
It’s a shame.
A CDN is not a storage service~ and even with a storage service you should have backups somewhere else (this is called "Disaster Recovery" - DR).
Ask Amazon AWS how many %%% is it currently in Bahrain and in the UAE (spoiler: S3 has been offline since the start of March 2026).
OP bought the CDN service, not the Storage service (my assumption from the title).
Bunny Storage lets you store and manage your files on bunny.net’s global network with fast performance and built-in replication. When you upload your files, bunny.net automatically copies them across regions so they are closer to your users. Your content stays available and loads quickly worldwide.ofc, bunny storage != s3, bunny invite to test s3 storage as a
previevfeature for nowNobody does it better 😈 naughtybunny
@daffyy My understanding is that you could use the Bunny CDN on the top of any storage service, or on the top of your own server (without the need of using Bunny Storage).
The OP will have to confirm if the Bunny Storage was used in one Single region, or if Replication was enabled to other Regions.
The Bunny documentation clearly states that: "To achieve eleven nines durability, enable at least one additional replication region beyond your primary region. Each additional replication region further reduces the chance of data loss".
You can't accuse the provider if you didn't want to pay for replication...
It was literally in the first post though:
Did you bother to read a single thing in this thread?
@ralf & @Obelous this is what is unclear to me: where is the data loss since the data was in another replication region?... If you get a 404 from one region, what stops you from making another call to the next (replication) region? https://docs.bunny.net/storage/http#storage-endpoints
=== Update: I have tested this and you can only connect to main/primary region when connecting by FTP/HTTP! However I understand that the 404 is coming from the CDN access, when linked to your Storage zone. So my guess is that the OP selected only 1x extra replication region (TBC).
@eran1243
1) were you using Bunny Storage? (not only the CDN)
2) which storage tier did you pick for your Bunny Storage service? (it is either HTTP/FTP or S3)
3) how many extra replications region(s) did you pick? (not counting your primary/main one)
And what exactly is laughable about that? It’s outright the best alternative to Cloudflare, and it’s European. Something that’s becoming more and more appreciated because of one specific person in the US, cough cough.
The funny thing is that Bunny’s CDN is actually faster than Cloudflare’s, and it also stayed online when Cloudflare decided to mess around for a bit and took down a percentage of the internet?
They have responded publicly in the meantime, which I think is fair. Here is their response:
It doesn’t seem like this is a general issue, and that this is on OP. Also, only complaining after 15 months? Sorry, but then it should have been reported earlier. This honestly just feels pathetic. I personally also use Bunny’s Storage, with replication enabled on both tiers, and I’ve had absolutely no issues at all.
That’s a feature of Bunny cdn
they should rebrand to lowendcdn