[SOLVED] S3 Hetzner Garbage collect fails with `error reading a body from connection`

[harunk]

Hello,

we have backup configured to Hetzner and the backup job itself is working without fail. The issue is with the garbage collection job that keeps failing most of the time. The logs are as follows

```
2026-05-12T02:30:01+02:00: Start GC phase1 (mark used chunks)
2026-05-12T02:30:01+02:00: marked 1% (8 of 758 index files)
2026-05-12T02:30:01+02:00: marked 2% (16 of 758 index files)
2026-05-12T02:30:01+02:00: marked 3% (23 of 758 index files)
2026-05-12T02:30:01+02:00: marked 4% (31 of 758 index files)
.
.
.
026-05-12T02:30:16+02:00: marked 98% (743 of 758 index files)
2026-05-12T02:30:16+02:00: marked 99% (751 of 758 index files)
2026-05-12T02:30:16+02:00: marked 100% (758 of 758 index files)
2026-05-12T02:30:16+02:00: Start GC phase2 (sweep unused chunks)
2026-05-12T02:31:37+02:00: queued notification (id=51d67b21-59d3-48d9-8097-96a3a1288546)
2026-05-12T02:31:37+02:00: TASK ERROR: error reading a body from connection: unexpected EOF during chunk size line

```

Basically the Phase2 fails as soon as it starts. Any pointers on what i could check?

Thanks!

 

[Chris]

Please do not double (or triple) post. Let's continue in https://forum.proxmox.com/threads/183483/

 

[harunk]

Sorry for the mixup! those posts were created yesterday but i didn't see them in my feed. Today morning i checked the forum again to ensure that these posts are not visible before creating a new one.

 

[fabian]

they were stuck in the moderation queue