"User has incomplete jobs running."

[erbader]

I'm trying out UCDeconvolve and was able to successfully run it on a couple of datasets, but when I tried to re-run some analyses, I am running into this error:

`2023-06-21 13:26:26,991|[UCD]|INFO: Starting UCDeconvolveBASE Run. | Timer Started.
Preprocessing Dataset | 100% (173 of 173) || Elapsed Time: 0:00:36 Time: 0:00:36
2023-06-21 13:27:08,402|[UCD]|ERROR: {"message":"User has incomplete jobs running."}
2023-06-21 13:27:08,402|[UCD]|ERROR: {"message":"User has incomplete jobs running."}
2023-06-21 13:27:08,405|[UCD]|INFO: Run Complete | Elapsed Time: 41.412 (s)

TypeError Traceback (most recent call last)
Cell In[7], line 1
----> 1 ucd.tl.base(adata)

File ~/anaconda3/envs/ucdenv/lib/python3.8/site-packages/ucdeconvolve/_tools/_base.py:135, in deconvolve_base(data, token, split, sort, propagate, return_results, key_added, use_raw, verbosity)
132 params = {"propagate" : propagate, "split" : split, "sort" : sort}
134 # Get signed URL
--> 135 metadata, url = ucdapi.start_run_request("base", params = params)
137 # Upload to URL
138 with ucdutils.timer(ucdlogger, "Uploading Data", "Upload Complete", logging.INFO):

File ~/anaconda3/envs/ucdenv/lib/python3.8/site-packages/ucdeconvolve/_api/_api.py:661, in start_run_request(runtype, params, token)
658 data = apiutils.process_response(response)
660 # Split response
--> 661 metadata = data['metadata']
662 url = data['url']
664 # Return split response

TypeError: string indices must be integers`

I was able to run the same code previously, so I don't think it's an issue with the code or input data. Can you please let me know how I can resolve this?

[dchary]

Hi @erbader thanks for noting this. We have cleared a logging cache that was causing this issue and it should be resolved. Additionally, we are working on a long-term patch to prevent this from reoccurring again. Please let us know if this error re-appears!

[madscientistca]

Ran into this annoying problem and did not see this post, but put it through the debugger and found it was being generated in the line:

response = apiutils.send_request("api/start_run", payload, token)

in the start_run_request call, where response was set to: "User has incomplete jobs running."

to address this without having to wait for others to clear the cache for a user, I modified the "init.py" file in the _api directory of the ucdeconvolve install in my virtual environment as follows:

from ._api import authenticate, register, activate, list_run_states, kill_run, remove_run

where I added the list_run_states, kill_run, and remove_run functions in the _api.py file, so they would be exposed. Now I run a check in my script to run ucdeconvolve by running list_run_states once I have authenticated by API key, and checked the 'completed' list to see if there are jobs that are indicated as not completed (that is, the completed data for a given run id is set to 'False', and then I kill and remove those. That fixes the problem.

[haochengyu2447]

Hi, I ran into the same problem. It showed "User has incomplete jobs running." at first, but after a few more attempts, it became "Invalid token".
I would really appreciate your help!

[francomich12]

hello. I am running into the same problem of "User has incomplete jobs running." i would also appreciate help!

[DaniHiam]

I was just wondering has there been any progress regarding this error? or is there a way for the user to clear their logging cache.