Re-run old job after upgrade

Troubleshooting
1

Hi all,

I prematurely cleared a patch motion correction job to clear up space some time ago, now unfortunately I have to continue refinement.
In between I’ve updated the cryosparc instance making it impossible to restart the job.

> [CPU: 264.8 MB]  Error occurred while processing J85/imported/20200824_134645_24495147_24494538.mrc.bz2
> Traceback (most recent call last):
>   File "/srv/public/cryosparc2/cryosparc2_worker/cryosparc_compute/jobs/pipeline.py", line 59, in exec
>     return self.process(item)
>   File "cryosparc_worker/cryosparc_compute/jobs/motioncorrection/run_patch.py", line 190, in cryosparc_compute.jobs.motioncorrection.run_patch.run_patch_motion_correction_multi.motionworker.process
>   File "cryosparc_worker/cryosparc_compute/jobs/motioncorrection/run_patch.py", line 193, in cryosparc_compute.jobs.motioncorrection.run_patch.run_patch_motion_correction_multi.motionworker.process
>   File "cryosparc_worker/cryosparc_compute/jobs/motioncorrection/run_patch.py", line 197, in cryosparc_compute.jobs.motioncorrection.run_patch.run_patch_motion_correction_multi.motionworker.process
>   File "/srv/public/cryosparc2/cryosparc2_worker/cryosparc_compute/util/__init__.py", line 50, in __getitem__
>     """%(key)
> AssertionError: "variable_dose" parameter not available. This job may be stale. 
>     Please try the following steps in order until the issue is resolved:
>     1. Build a new job and try again.
>     2. Run the following via command line: cryosparcm cli "refresh_job_types()".
>     3. Ensure no orphaned cryoSPARC processes are running in the background.
>     
> 
> Marking J85/imported/20200824_134645_24495147_24494538.mrc.bz2 as incomplete and continuing...

Any advice how to re-start this job while keeping the history?

Best,
Tarek

2

@team, any ideas hot to re-introduce motion correction?

3

pushing this again as I repeatedly face that problem.

4

Hi @tarek,

As per the error message, have you tried the following steps?

Please try the following steps in order until the issue is resolved:
1. Build a new job and try again.
2. Run the following via command line: cryosparcm cli "refresh_job_types()".
3. Ensure no orphaned cryoSPARC processes are running in the background.

5

Hi @sdawood,

the problem is that this job was run at the beginning of a long cascade of down-stream processes.
How can I re-feed or re-link the output of a new job at this very early position?
In particular I want to run local motion correction of “final” particles but global frame alignment information is missing. Just re-running the original patch motion correction doesn’t work as you can see from the log.

Suggestions 2 and 3 don’t solve the situation either.

Best,
Tarek

6

Hi @tarek,

Try creating a new Patch Motion job and connect the same inputs as the one you are trying to rerun. This should work; you don’t need to reprocess any of the previous jobs.

7

Hi @sdawood,

thanks for replying.
How can I connect the output of that new job as input for an already completed job?
If I simply create a new patch motion correction job, the aligned micrographs have no particle coordinate information. Previously extracted and refined particles have in turn no information of the frame alignment by the “new Patch motion job”.

Best,
Tarek