github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #1731] pulling manifest Error: EOF when pulling after disk is full #26747

New Issue
Open
opened 2026-04-22 03:16:07 -05:00 by GiteaMirror · 10 comments
GiteaMirror commented 2026-04-22 03:16:07 -05:00
Owner
Copy Link

Originally created by @jmorganca on GitHub (Dec 28, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/1731

Originally assigned to: @mxyng on GitHub.

To reproduce, pull with little disk space lefT:

$ ollama run deepseek-coder:33b
pulling manifest 
Error: write /usr/share/ollama/.ollama/models/blobs/sha256:065b9a7416ba28634cd4efc2cd3024d4755731c1275dc0286b81b01793185fbb-partial-0: no space left on device

Even with more space, future ollama pull commands fail until Ollama restarted:

$ ollama run deepseek-coder:33b
pulling manifest 
Error: EOF
Originally created by @jmorganca on GitHub (Dec 28, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/1731 Originally assigned to: @mxyng on GitHub. To reproduce, pull with little disk space lefT: ``` $ ollama run deepseek-coder:33b pulling manifest Error: write /usr/share/ollama/.ollama/models/blobs/sha256:065b9a7416ba28634cd4efc2cd3024d4755731c1275dc0286b81b01793185fbb-partial-0: no space left on device ``` Even with more space, future `ollama pull` commands fail until Ollama restarted: ``` $ ollama run deepseek-coder:33b pulling manifest Error: EOF ```
GiteaMirror added the bug label 2026-04-22 03:16:07 -05:00
GiteaMirror commented 2026-04-22 03:16:08 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Jan 16, 2024):

I haven't been able to reproduce this. Ollama behaves as expected. Here is the test I ran:

  1. Create a VM with multiple, small physical volumes
  2. Create a logical volume by attaching one of the physical volumes
  3. Format, mount, and configure as OLLAMA_MODELS
  4. Pull a large model, e.g. llama2:70b
  5. Step 4 should fail once the logical volume is full
  6. Expand the disk by attaching another physical volumes to the logical volume
  7. Repeat step 4 which should resume the download where it left off

@jmorganca do you have any more details on how to reproduce this?

<!-- gh-comment-id:1894285486 --> @mxyng commented on GitHub (Jan 16, 2024): I haven't been able to reproduce this. Ollama behaves as expected. Here is the test I ran: 1. Create a VM with multiple, small physical volumes 2. Create a logical volume by attaching _one_ of the physical volumes 3. Format, mount, and configure as `OLLAMA_MODELS` 4. Pull a large model, e.g. llama2:70b 5. Step 4 should fail once the logical volume is full 6. Expand the disk by attaching another physical volumes to the logical volume 7. Repeat step 4 which should resume the download where it left off @jmorganca do you have any more details on how to reproduce this?
GiteaMirror commented 2026-04-22 03:16:08 -05:00
Author
Owner
Copy Link

@Pytness commented on GitHub (Jan 31, 2024):

I just managed to unintentionally reproduce this bug.
The steps I followed:

  • Pull multiple models until disk is full.
  • Get disk is full error pulling another model.
  • ollama rm a few models.
  • Try to ollama pull a model and get EOF error.

Restarting ollama fixed the issue.

<!-- gh-comment-id:1919141927 --> @Pytness commented on GitHub (Jan 31, 2024): I just managed to unintentionally reproduce this bug. The steps I followed: - Pull multiple models until disk is full. - Get disk is full error pulling another model. - `ollama rm` a few models. - Try to `ollama pull` a model and get `EOF` error. Restarting ollama fixed the issue.
GiteaMirror commented 2026-04-22 03:16:09 -05:00
Author
Owner
Copy Link

@wg1k commented on GitHub (May 14, 2024):

> ollama --version
ollama version is 0.1.34

Restarting didn't work for me.

After having a look at https://github.com/ollama/ollama/pull/344, I ended with the below command:

sudo rm /usr/share/ollama/.ollama/models/blobs/*-partial*
<!-- gh-comment-id:2109943047 --> @wg1k commented on GitHub (May 14, 2024): ```shell-session > ollama --version ollama version is 0.1.34 ``` Restarting didn't work for me. After having a look at https://github.com/ollama/ollama/pull/344, I ended with the below command: ```shell sudo rm /usr/share/ollama/.ollama/models/blobs/*-partial* ```
GiteaMirror commented 2026-04-22 03:16:09 -05:00
Author
Owner
Copy Link

@candera commented on GitHub (Jul 23, 2024):

Deleting only the ones of zero size in ~/.ollama/models/blobs/ worked for me, and I didn't have to restart the entire download.

<!-- gh-comment-id:2246072799 --> @candera commented on GitHub (Jul 23, 2024): Deleting only the ones of zero size in `~/.ollama/models/blobs/` worked for me, and I didn't have to restart the entire download.
GiteaMirror commented 2026-04-22 03:16:09 -05:00
Author
Owner
Copy Link

@sezerogras commented on GitHub (Jan 28, 2025):

I had the same problem :
disk check macos : df -h
to clear memory : sudo purge
ollama rm (model name)
ollama run (new model name)
Restarting ollama fixes the issue.

<!-- gh-comment-id:2618893898 --> @sezerogras commented on GitHub (Jan 28, 2025): I had the same problem : disk check macos : df -h to clear memory : sudo purge ollama rm (model name) ollama run (new model name) Restarting ollama fixes the issue.
GiteaMirror commented 2026-04-22 03:16:10 -05:00
Author
Owner
Copy Link

@DarkTyger commented on GitHub (Apr 7, 2025):

Confirmed.

$ ollama --version
ollama version is 0.6.4
$ ./outline.sh 
pulling manifest 
Error: EOF

The fix was to restart ollama serve. Seems ollama is incorrectly caching the state of the disk, even after issuing ollama rm to clear up space.

<!-- gh-comment-id:2782206202 --> @DarkTyger commented on GitHub (Apr 7, 2025): Confirmed. ``` bash $ ollama --version ollama version is 0.6.4 $ ./outline.sh pulling manifest Error: EOF ``` The fix was to restart `ollama serve`. Seems ollama is incorrectly caching the state of the disk, even after issuing `ollama rm` to clear up space.
GiteaMirror commented 2026-04-22 03:16:10 -05:00
Author
Owner
Copy Link

@yoliverasPozo commented on GitHub (Apr 9, 2025):

restarting ollama worked for me, thanks. Now I only have to work out why ollama keeps hanging while pulling models so I don't have to use the timeout shell script.

<!-- gh-comment-id:2789987343 --> @yoliverasPozo commented on GitHub (Apr 9, 2025): restarting ollama worked for me, thanks. Now I only have to work out why ollama keeps hanging while pulling models so I don't have to use the timeout shell script.
GiteaMirror commented 2026-04-22 03:16:11 -05:00
Author
Owner
Copy Link

@Nyx1197 commented on GitHub (May 15, 2025):

Deleting only the ones of zero size in ~/.ollama/models/blobs/ worked for me, and I didn't have to restart the entire download.删除 ~/.ollama/models/blobs/ 目录中零大小的文件，这对我有用，并且不需要重启整个下载。

thanks, that help a lot.

<!-- gh-comment-id:2882156792 --> @Nyx1197 commented on GitHub (May 15, 2025): > Deleting only the ones of zero size in `~/.ollama/models/blobs/` worked for me, and I didn't have to restart the entire download.删除 `~/.ollama/models/blobs/` 目录中零大小的文件，这对我有用，并且不需要重启整个下载。 thanks, that help a lot.
GiteaMirror commented 2026-04-22 03:16:12 -05:00
Author
Owner
Copy Link

@debuglevel commented on GitHub (Nov 24, 2025):

Would be nice to see this fixed. This is an issue which is rather hard to handle automatically.

<!-- gh-comment-id:3571148852 --> @debuglevel commented on GitHub (Nov 24, 2025): Would be nice to see this fixed. This is an issue which is rather hard to handle automatically.
GiteaMirror commented 2026-04-22 03:16:13 -05:00
Author
Owner
Copy Link

@bysiber commented on GitHub (Mar 5, 2026):

This is a painful one, once Ollama hits a disk-full state during a pull, the partial download can leave things in a broken state where even retrying with more space fails until you restart.

A few things that help recover and prevent this:

# 1. Check current disk space
df -h /

# 2. See how much space Ollama models are using
du -sh ~/.ollama/models/

# 3. Remove partial/stuck downloads
rm -rf ~/.ollama/models/blobs/*-partial-*

# 4. Restart Ollama after cleanup
pkill ollama && ollama serve &

# 5. Retry the pull
ollama pull deepseek-coder:33b

To avoid running out of space in the first place, keeping dev caches clean helps a lot. Tools like Docker, Xcode DerivedData, node_modules, and Homebrew can silently eat 20-50GB on macOS.

I built ClearDisk — a free, open-source macOS menu bar utility that monitors 44+ cache paths and shows which ones are eating your disk. Helps keep enough headroom for large model downloads.

<!-- gh-comment-id:4001282217 --> @bysiber commented on GitHub (Mar 5, 2026): This is a painful one, once Ollama hits a disk-full state during a pull, the partial download can leave things in a broken state where even retrying with more space fails until you restart. A few things that help recover and prevent this: ```bash # 1. Check current disk space df -h / # 2. See how much space Ollama models are using du -sh ~/.ollama/models/ # 3. Remove partial/stuck downloads rm -rf ~/.ollama/models/blobs/*-partial-* # 4. Restart Ollama after cleanup pkill ollama && ollama serve & # 5. Retry the pull ollama pull deepseek-coder:33b ``` To avoid running out of space in the first place, keeping dev caches clean helps a lot. Tools like Docker, Xcode DerivedData, node_modules, and Homebrew can silently eat 20-50GB on macOS. I built [ClearDisk](https://github.com/bysiber/cleardisk) — a free, open-source macOS menu bar utility that monitors 44+ cache paths and shows which ones are eating your disk. Helps keep enough headroom for large model downloads.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#26747
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?