Segfault on query

[mjvmroz]

I have bor running in Docker via Odin's docker-compose configuration, and the server starts fine, but then segfaults when the Obsidian plugin tries to make queries:

2023-09-25 13:05:29 INFO:     Started server process [30]
2023-09-25 13:05:29 INFO:     Waiting for application startup.
2023-09-25 13:05:29 INFO:     Application startup complete.
2023-09-25 13:05:29 INFO:     Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
2023-09-25 13:05:42 INFO:     172.18.0.1:52014 - "OPTIONS /knowledge_base/general/get_all_for_repo HTTP/1.1" 200 OK
2023-09-25 13:05:42 INFO:     172.18.0.1:52020 - "OPTIONS /knowledge_base/general/get_all_for_repo HTTP/1.1" 200 OK
2023-09-25 13:05:42 INFO:     172.18.0.1:52014 - "POST /knowledge_base/general/get_all_for_repo HTTP/1.1" 204 No Content
2023-09-25 13:05:42 INFO:     172.18.0.1:52020 - "POST /knowledge_base/general/get_all_for_repo HTTP/1.1" 204 No Content
2023-09-25 13:05:43 qemu: uncaught target signal 11 (Segmentation fault) - core dumped
2023-09-25 13:05:43 /tmp/tmpalqwjtx0: line 3:    30 Segmentation fault      uvicorn core.restapi.api:app --host 0.0.0.0 --port 8000
2023-09-25 13:05:43 ERROR conda.cli.main_run:execute(47): `conda run uvicorn core.restapi.api:app --host 0.0.0.0 --port 8000` failed. (See above for error)

I've played around a bit and I don't think it's just one resource causing problems, but an example of a request which seems to trigger crashes is:

POST http://localhost:8000/knowledge_base/notes/update_file

{
  "path":"/Users/mroz/Library/Mobile Documents/iCloud~md~obsidian/Documents/Main/Untitled.md",
  "type":"Notes",
  "content":""
}

[trunksio]

I am seeing this issue as well: bor | INFO: 172.21.0.1:41960 - "OPTIONS /knowledge_base/notes/update_file HTTP/1.1" 200 OK bor | INFO: 172.21.0.1:41970 - "POST /knowledge_base/general/get_all_for_repo HTTP/1.1" 204 No Content bor | INFO: 172.21.0.1:41978 - "POST /knowledge_base/general/get_all_for_repo HTTP/1.1" 204 No Content bor | qemu: uncaught target signal 11 (Segmentation fault) - core dumped bor | /tmp/tmpyptbg8p6: line 3: 30 Segmentation fault uvicorn core.restapi.api:app --host 0.0.0.0 --port 8000 bor | ERROR conda.cli.main_run:execute(47): conda run uvicorn core.restapi.api:app --host 0.0.0.0 --port 8000 failed. (See above for error) bor exited with code 139

[xc0f0c0bx]

same here. what is this? seems abandon

[katarinasupe]

Hi all, thank you for opening and commenting on the issue. I will try to get more information from the project contributors. The reason for no one answering sooner is that the main contributor is not active here anymore, so we'll have to check what the issue is internally. If you have any idea on how to resolve this, we are open to contributions 😄

[AlexIchenskiy]

It's a known issue with ODIN, along with bugs in some other endpoints, which is being addressed here: memgraph/odin#8.