How to override timout for msa-server?

Expected Behavior

Jobs do not fail after one hour

Current Behavior

Jobs fail after a timeout of one hour

Steps to Reproduce (for bugs)

For larger sequences msa-server jobs fail after one hour, mostly limited by NFS disk I/O. Rerunning helps only in some cases.

ColabFold Output (for bugs)

"Timeout" in server's output, generic error on colabfold_batch's output.
ERROR: | | 3585/? [elapsed: 1:00:01 remaining: 00:00]
2022-10-27 14:39:17,529 Could not get MSA/templates for 4D2I_1: MMseqs2 API is giving errors. Please confirm your input is a valid protein sequence. If error persists, please try again an hour later.
Traceback (most recent call last):
File "batch.py", line 1333, in run
) = get_msa_and_templates(
File "batch.py", line 820, in get_msa_and_templates
a3m_lines = run_mmseqs2(
File "colabfold.py", line 199, in run_mmseqs2
raise Exception(f'MMseqs2 API is giving errors. Please confirm your input is a valid protein sequence. If error persists, please try again an hour later.')
Exception: MMseqs2 API is giving errors. Please confirm your input is a valid protein sequence. If error persists, please try again an hour later.

Context

colabfold_batch --host-url=http://our.local.server:8888 <input_dir> <output_dir>

Your Environment

Shared application server running RHEL7-9.0, 512G RAM (not for exclusive colabfold use), remote networked file systems (NFS/gpfs)
MMSEQS_COMMIT=${1:-4589151554eb83a70ff0c4d04d21b83cabc203e4}
BACKEND_COMMIT=${2:-14e087560f309f989a5e1feb54fd1f9c988076d5}

Review the run_mmseqs2 path in colabfold.py and its caller in batch.py, then inspect the msa-server configuration used with --host-url. Reproduce a long job in the stated NFS environment and identify where the one-hour timeout is applied. Done means the timeout can be overridden and larger jobs complete without the generic server error.