Fizzled firework cannot rerun

FireWorks
1

Hi there,

I have a firework with fw_id of 14980 fizzled. I can run lpad get_fws -i 14980 and successfully get its status:

{
    "fw_id": 14980,
    "created_on": "2024-02-10T05:05:40.673328",
    "updated_on": "2024-02-27T20:45:43.873777",
    "state": "FIZZLED",
    "name": "store_inputs"
}

However, when I try to lpad rerun_fws -i 14980, it failed with the error:

Traceback (most recent call last):
  File "/dartfs-hpc/rc/home/x/anaconda3/envs/ht_defects/bin/lpad", line 8, in <module>
    sys.exit(lpad())
  File "/dartfs-hpc/rc/home/x/anaconda3/envs/ht_defects/lib/python3.10/site-packages/fireworks/scripts/lpad_run.py", line 1551, in lpad
    args.func(args)
  File "/dartfs-hpc/rc/home/x/anaconda3/envs/ht_defects/lib/python3.10/site-packages/fireworks/scripts/lpad_run.py", line 641, in rerun_fws
    lp.rerun_fw(int(f), recover_launch=l, recover_mode=args.recover_mode)
  File "/dartfs-hpc/rc/home/x/anaconda3/envs/ht_defects/lib/python3.10/site-packages/fireworks/core/launchpad.py", line 1714, in rerun_fw
    with WFLock(self, fw_id):
  File "/dartfs-hpc/rc/home/x/anaconda3/envs/ht_defects/lib/python3.10/site-packages/fireworks/core/launchpad.py", line 129, in __enter__
    raise ValueError(f"Could not find workflow in database: {self.fw_id}")
ValueError: Could not find workflow in database: 14980

Does any have a clue of this? Thanks in advance!

2

Hi @Zhenkun_Yuan, you might have a lock on your firework that is preventing it from being rerun. Try running:

lpad admin unlock -i 14980

before trying to rerun the firework again.

3

Thanks a lot! @Aaron_Kaplan, though still get the same error Hmm

4

Is this firework the only one in the workflow? If there are other fireworks in the flow, or they fizzled, or are stuck (run lpad detect_lostruns to check), there might be other issues at hand

5

Are you able to find the workflow corresponding to the FW?

lpad get_wfs -i 14980

6

@Anubhav_Jain @Aaron_Kaplan Thanks for helping. It looks the workflow associated with this fw_id has been lost, as I ran lpad get_wfs -i 14980 or other fw_id, the return is [] (i.e., empty). I checked other workflows, they are no problems.

The lpad detect_lostruns ran into TypeError: 'NoneType' object is not subscriptable

Since the calculations are not heavy, I will just restart the whole workflow.