--fail-percent not working when using server.shell

[avrilmaomao]

Describe the bug

In the pyinfra cli, I added the parameters " --serial" and "--fail-percent 0" to let the deploy run one host at a time , and if one host encountered error, the deploy process would stop.
When I use "exec" to execute adhoc command, the --fail-percent worked as expected.But when I switch to a .py file which using "server.shell" to execute the same command, the deploy was still executed on all hosts even the first host had failed.

To Reproduce

• Operation code & usage

1. write a simple deploy file(eg. test_fail_percent.py), using server.shell to execute a command that does not exist.

import pyinfra.operations.server as server

server.shell('uptim')

2. run pyinfra with the above file with multiple hosts
   pyinfra -y --serial --fail-percent 0 --debug host1,host2 test_fail_percent.py

• Target system information
  ubuntu 22.04

Expected behavior

The deploy should be stopped for other hosts when the first host encountered error during the "server.shell" execution

Meta

• Include output of pyinfra --support.
  System: Linux
  Platform: Linux-5.15.0-101-generic-x86_64-with-glibc2.35
  Release: 5.15.0-101-generic
  Machine: x86_64
  pyinfra: v3.1
  click: v8.1.7
  configparser: v7.0.0
  distro: v1.9.0
  gevent: v24.2.1
  jinja2: v3.1.3
  packaging: v24.1
  paramiko: v3.4.0
  python-dateutil: v2.9.0.post0
  pywinrm: v0.4.3
  setuptools: v74.1.1
  typeguard: v4.3.0
  typing-extensions: v4.12.2
  wheel: v0.44.0
  Executable: /usr/sbin/pyinfra
  Python: 3.10.12 (CPython, GCC 11.4.0)
• How was pyinfra installed (source/pip)?
  pip
• Consider including output with -vv and --debug.

-> Loading config...
--> Loading inventory...
[pyinfra_cli.inventory] Creating fake inventory...
[pyinfra_cli.inventory] Checking possible group_data at: /var/lib/jenkins/group_data

--> Connecting to hosts...
[pyinfra.connectors.ssh] Connecting to: host1 ({'allow_agent': True, 'look_for_keys': True, '_pyinfra_ssh_forward_agent': False, '_pyinfra_ssh_config_file': None, '_pyinfra_ssh_known_hosts_file': None, '_pyinfra_ssh_strict_host_key_checking': 'accept-new', '_pyinfra_ssh_paramiko_connect_kwargs': None, 'timeout': 10})
[pyinfra.connectors.sshuserclient.client] Loading SSH config: None
[pyinfra.connectors.ssh] Connecting to: host2 ({'allow_agent': True, 'look_for_keys': True, '_pyinfra_ssh_forward_agent': False, '_pyinfra_ssh_config_file': None, '_pyinfra_ssh_known_hosts_file': None, '_pyinfra_ssh_strict_host_key_checking': 'accept-new', '_pyinfra_ssh_paramiko_connect_kwargs': None, 'timeout': 10})
[host1] Connected
[host2] Connected
[pyinfra.api.state] Activating host: host1
[pyinfra.api.state] Activating host: host2

--> Preparing operations...
--> Preparing Operations...
Loading: test_op.py
[pyinfra.api.operation] Adding operation, {'server.shell'}, opOrder=(0, 3), opHash=cedda09f9ebcba03bb3a36b21a0111fc8b7c3ec0
[host2] Ready: test_op.py
[pyinfra.api.operation] Adding operation, {'server.shell'}, opOrder=(0, 3), opHash=cedda09f9ebcba03bb3a36b21a0111fc8b7c3ec0
[host1] Ready: test_op.py

--> Skipping change detection
--> Beginning operation run...
[pyinfra.api.operations] Running all ops on host2

--> Starting operation: server.shell (uptim)
[pyinfra.api.operations] Starting operation {'server.shell'} on host2
[pyinfra.connectors.ssh] Running command on host2: (pty=False) sh -c uptim
[pyinfra.connectors.ssh] Waiting for exit status...
[pyinfra.connectors.ssh] Command exit status: 127
[host2] sh: 1: uptim: not found
[host2] Error: executed 0 commands
[pyinfra.api.state] Failing hosts: host2
[pyinfra.api.operations] Running all ops on host1

--> Starting operation: server.shell (uptim)
[pyinfra.api.operations] Starting operation {'server.shell'} on host1
[pyinfra.connectors.ssh] Running command on host1: (pty=False) sh -c uptim
[pyinfra.connectors.ssh] Waiting for exit status...
[pyinfra.connectors.ssh] Command exit status: 127
[host1] sh: 1: uptim: not found
[host1] Error: executed 0 commands
[pyinfra.api.state] Failing hosts: host1

[avrilmaomao]

Hi team,
Is there a plan to fix this or any way to get around it?