Rustic hangs after about 400GiB of checked data

[3tes]

I am using rustic to backup my whole image library wich is around 700 GiB. For the backend I am using rclone with sftp and a hetzner-storage box. This is my current config:

repository = "rclone:hetzner-sb-1:PATH"
password = "PASSWORD"
cache-dir = "/mnt/photos/cache"

[backup.hooks]
run-after = [
    "rustic forget",
     "curl -m 10 --retry 5 UPTIMECUMA"
    ]
run-failed = ["curl -m 10 --retry 5 UPTIMECUMA"]

[[backup.snapshots]]
sources = [
    "/mnt/photos/immich",
    "/mnt/photos/external_libraries"
    ]

[forget]
prune = true
keep-daily = 31
keep-monthly = 12
keep-yearly = 3

[global.env]
RCLONE_CHECKERS = "4"
RCLONE_TRANSFERS = "4"

And this is my rclone config:

[hetzner-sb-1]
type = sftp
host = USERNAME.your-storagebox.de
user = USERNAME
port = 23
key_file = ~/.ssh/hetzner-sb-1
shell_type = unix
md5sum_command = md5sum
sha1sum_command = sha1sum
known_hosts_file = ~/.ssh/hetzner_known_hosts

Up until now everything was working fine but if I run rustic backup now it runs normal until it reaches about 400 GiB and the it stalls/hangs and does nothing. I enabled the trace output-log and it shows that the last line is repeated indefinitely:

2025-06-29T11:09:20.010612374Z [DEBUG] (1) rustic_core::archiver::tree_archiver: unchanged tree: DIRECTORY
2025-06-29T11:09:20.010953943Z [DEBUG] (1) rustic_core::archiver::tree_archiver: unchanged file: JPG FILE
2025-06-29T11:10:10.171758181Z [TRACE] (4) hyper_util::client::legacy::pool: [/home/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/hyper-util-0.1.14/src/client/legacy/pool.rs:808] idle interval checking for expired
2025-06-29T11:11:40.170637527Z [TRACE] (4) hyper_util::client::legacy::pool: [/home/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/hyper-util-0.1.14/src/client/legacy/pool.rs:808] idle interval checking for expired
2025-06-29T11:13:10.17082878Z [TRACE] (4) hyper_util::client::legacy::pool: [/home/runner/.cargo/registry/src/index.crates.io-1949cf8c6b5b557f/hyper-util-0.1.14/src/client/legacy/pool.rs:808] idle interval checking for expired

I also ran the command twice and it stalls after the exact same files. What can I do to debug issue further or is this a known problem ?

[SomeThink1729]

Have you tried isolating the issue to the network connection?

If possible try to backup to a repo on the same device and see if the issue remains

[3tes]

I find it unlikely that it is a network problem. Because the progress always stops at the exact same file and if I remove that folder the backup works fine.