Do not init disabled plugins (need discussion)

[AnthonyBe]

Describe the bug
On Glances 4.0.4 start-up, and the following error is logged:
ERROR -- containers (Docker) plugin - Can't connect to Docker (Error while fetching server API version: (2, 'CreateFile', 'The system cannot find the file specified.'))
even though the containers plugin is disabled.
Glances continues to run without issue so low priority issue. It just seems odd that with the containers plugin disabled, it is obviously still trying to load something. Or is it just trying to detect whenther it is running within a Docker container on start-up and this is how it logs that failed detection?

To Reproduce
Steps to reproduce the behavior:

1. Start Glances with the following options glances -w --fs-free-space --enable-process-extended --config %APPDATA%\glances\glances.conf
2. glances.conf contains:

[containers]
disable=True

4. The error is logged on start-up only.

2024-05-16 13:54:16,477 -- WARNING -- PsUtil can not grab processes gids ('Process' object has no attribute 'gids')
2024-05-16 13:54:16,477 -- INFO -- Start Glances 4.0.4
2024-05-16 13:54:16,493 -- INFO -- CPython 3.12.3 (C:\Python312\python.exe) and psutil 5.9.8 detected
2024-05-16 13:54:16,524 -- INFO -- Read configuration file 'C:\Users\Administrator\AppData\Roaming\glances\glances.conf'
2024-05-16 13:54:19,146 -- INFO -- Start GlancesWebServer mode
2024-05-16 13:54:20,292 -- ERROR -- containers (Docker) plugin - Can't connect to Docker (Error while fetching server API version: (2, 'CreateFile', 'The system cannot find the file specified.'))
2024-05-16 13:54:20,481 -- ERROR -- Can not init battery class ([WinError 3] The system cannot find the path specified: '/sys/class/power_supply')
2024-05-16 13:54:24,329 -- INFO -- Glances RESTful API Server started on http://0.0.0.0:61208/api/4
2024-05-16 13:54:24,329 -- INFO -- Get WebUI in C:\Python312\Lib\site-packages\glances\outputs\static/public
2024-05-16 13:54:24,329 -- INFO -- Glances Web User Interface started on http://0.0.0.0:61208/

Expected behavior
no containers code loading and therefore no error logged

Screenshots
If applicable, add screenshots to help explain your problem.

Environment (please complete the following information)

• Operating System (lsb_release -a or OS name/version): Windows Server 2016
• Glances & psutil versions: Glances version: 4.0.4, PsUtil version: 5.9.8
• How do you install Glances (Pypi package, script, package manager, source): Pypi package
• Glances test (only available with Glances 3.1.7 or higher):

To be completed with result of: glances --issue
C:\WINDOWS\system32>glances --issue
===============================================================================
Glances �[94m4.0.4�[0m (C:\Python312\Lib\site-packages\glances\__init__.py)
Python �[94m3.12.3�[0m (C:\Python312\python.exe)
PsUtil �[94m5.9.8�[0m (C:\Python312\Lib\site-packages\psutil\__init__.py)
===============================================================================
�[94malert         �[92m[OK]   �[94m 0.00000s
�[0m[]�[0m
�[94mamps          �[92m[OK]   �[94m 0.00000s �[93mkey=name �[0m
[{'count': 0,
 'countmax': None,
 'countmin': 1.0,
 'key': 'name',
 'name': 'Dropbox',
 'refresh': 3.0,
 'regex': True,
 'result': None,
 'timer': 5.043111801147461}, ...]�[0m
�[94mcloud�[0m         [NA]�[0m�[0m
�[94mconnections�[0m   [NA]�[0m�[0m
�[94mcontainers�[0m    [NA]�[0m�[0m
�[94mcore          �[92m[OK]   �[94m 0.00000s
�[0m{'log': 2, 'phys': 2}�[0m
�[94mcpu           �[92m[OK]   �[94m 0.00000s
�[0m{'cpucore': 2,
'ctx_switches': 3188,
'ctx_switches_gauge': 9947357,
'ctx_switches_rate_per_sec': 1577.0,
'dpc': 0.0,
'idle': 97.7,
'interrupts': 3055,
'interrupts_gauge': 10091286,
'interrupts_rate_per_sec': 1511.0,
'soft_interrupts': 0,
'soft_interrupts_gauge': 0,
'soft_interrupts_rate_per_sec': 0.0,
'syscalls': 8936,
'syscalls_gauge': 73725888,
'syscalls_rate_per_sec': 4421.0,
'system': 0.8,
'time_since_update': 2.021165370941162,
'total': 30.3,
'user': 1.6}�[0m
�[94mdiskio        �[92m[OK]   �[94m 0.00000s �[93mkey=disk_name �[0m
[{'disk_name': 'PhysicalDrive0',
 'key': 'disk_name',
 'read_bytes': 692224,
 'read_bytes_gauge': 7445936128,
 'read_bytes_rate_per_sec': 342487.0,
 'read_count': 139,
 'read_count_gauge': 288206,
 'read_count_rate_per_sec': 68.0,
 'time_since_update': 2.021165370941162,
 'write_bytes': 1449984,
 'write_bytes_gauge': 6838657536,
 'write_bytes_rate_per_sec': 717399.0,
 'write_count': 178,
 'write_count_gauge': 470832,
 'write_count_rate_per_sec': 88.0}, ...]�[0m
�[94mfolders�[0m       [NA]�[0m�[0m
�[94mfs            �[92m[OK]   �[94m 0.01559s �[93mkey=mnt_point �[0m
[{'device_name': 'C:\\',
 'free': 213466804224,
 'fs_type': 'NTFS',
 'key': 'mnt_point',
 'mnt_point': 'C:\\',
 'percent': 61.1,
 'size': 549048020992,
 'used': 335581216768}, ...]�[0m
�[94mgpu�[0m           [NA]�[0m�[0m
�[94mhelp          �[92m[OK]   �[94m 0.00000s
�[0mNone�[0m
�[94mip            �[92m[OK]   �[94m 2.03416s
�[0m{'address': '***', 'gateway': '***', 'mask': '***', 'mask_cidr': '***'}�[0m
�[94mirq�[0m           [NA]�[0m�[0m
�[94mload          �[92m[OK]   �[94m 0.00000s
�[0m{'cpucore': 2, 'min1': 0.08, 'min15': 0.01, 'min5': 0.02}�[0m
�[94mmem           �[92m[OK]   �[94m 0.00000s
�[0m{'available': 7180079104, 'free': 7180079104, 'percent': 58.1, 'total': 17143681024, 'used': 9963601920}�[0m
�[94mmemswap       �[92m[OK]   �[94m 0.00000s
�[0m{'free': 2296844288,
'percent': 9.9,
'sin': 0,
'sout': 0,
'time_since_update': 5.98190450668335,
'total': 2550136832,
'used': 253292544}�[0m
�[94mnetwork       �[92m[OK]   �[94m 0.01563s �[93mkey=interface_name �[0m
[{'alias': None,
 'bytes_all': 0,
 'bytes_all_gauge': 0,
 'bytes_all_rate_per_sec': 0.0,
 'bytes_recv': 0,
 'bytes_recv_gauge': 0,
 'bytes_recv_rate_per_sec': 0.0,
 'bytes_sent': 0,
 'bytes_sent_gauge': 0,
 'bytes_sent_rate_per_sec': 0.0,
 'interface_name': 'Ethernet 2',
 'key': 'interface_name',
 'speed': 0,
 'time_since_update': 4.07091760635376}, ...]�[0m
�[94mnow           �[92m[OK]   �[94m 0.00000s
�[0m{'custom': '2024-05-16 14:02:46 +1000', 'iso': '2024-05-16T14:02:46+10:00'}�[0m
�[94mpercpu        �[92m[OK]   �[94m 0.00000s �[93mkey=cpu_number �[0m
[{'cpu_number': 0, 'idle': 63.1, 'key': 'cpu_number', 'system': 8.4, 'total': 36.9, 'user': 26.7}, ...]�[0m
�[94mports         �[92m[OK]   �[94m 0.00000s �[93mkey=indice �[0m
[{'description': 'DefaultGateway',
 'host': '192.168.0.1',
 'indice': 'port_0',
 'key': 'indice',
 'port': 0,
 'refresh': 30,
 'rtt_warning': None,
 'status': 0.126356,
 'timeout': 3}, ...]�[0m
�[94mprocesscount  �[92m[OK]   �[94m 0.12500s
�[0m{'pid_max': 0, 'running': 83, 'sleeping': 0, 'thread': 1013, 'total': 85}�[0m
�[94mprocesslist   �[92m[OK]   �[94m 0.00000s
�[0m[]�[0m
�[94mpsutilversion �[92m[OK]   �[94m 0.00000s
�[0m'5.9.8'�[0m
�[94mquicklook     �[92m[OK]   �[94m 0.00000s
�[0m{'cpu': 30.3,
'cpu_hz': 2500000000.0,
'cpu_hz_current': 2125000000.0,
'cpu_log_core': 2,
'cpu_name': 'CPU',
'cpu_phys_core': 2,
'load': 0.5,
'mem': 58.1,
'percpu': [{...}, {...}],
'swap': 9.9}�[0m
�[94mraid�[0m          [NA]�[0m�[0m
�[94msensors�[0m       [NA]�[0m�[0m
�[94msmart�[0m         [NA]�[0m�[0m
�[94msystem        �[92m[OK]   �[94m 0.00000s
�[0m{'hostname': 'ANTMICROSERVER',
'hr_name': 'Windows 2016Server SP0 64bit',
'os_name': 'Windows',
'os_version': '2016Server SP0',
'platform': '64bit'}�[0m
�[94muptime        �[92m[OK]   �[94m 0.00000s
�[0m{'seconds': 6831}�[0m
�[94mversion       �[92m[OK]   �[94m 0.00000s
�[0m'4.0.4'�[0m
�[94mwifi�[0m          [NA]�[0m�[0m
===============================================================================
Total time to update all stats: �[94m2.19037s�[0m
===============================================================================

Additional context
n/a

You can also pastebin:

• the Glances configuration file
  glances.conf.txt

• the Glances log file (https://glances.readthedocs.io/en/latest/config.html#logging)
  glances-Administrator.log

[nicolargo]

For the moment all plugins are initialized (even if they are disabled in the configuration file).

Only enabled plugins are updated.

The error is raised durin the initialisation.

[AnthonyBe]

For the moment all plugins are initialized (even if they are disabled in the configuration file).

I'm guessing that means that further down the track, as code is further optimized, it would be possible to only initialize enabled plugins which would (in theory) also speed start up, reduce memory footprint, etc.?

[nicolargo]

Correct but a "disabled" plugin can be enable during execution (for example by pressing 'D' hotkey for the containers plugin). So in this case, the plugin should be initialized in the loop...

We have to think about it, but the proposal is interresting.

cc: @RazCrimson

[AnthonyBe]

Ahh... yep, OK.

[RazCrimson]

Currently, all plugins are initialized irrespective of them being disabled or enabled.

Instead of initializing at start, maybe we could dynamically attempt a initialization when the plugin gets enabled.
Similarly, when a plugin gets disabled, we stop the plugin and clean out the plugin from memory too?

This would make show/hide on the screen quite more laggy/slow than it currently is.
But this should reduce the launch time of glances as well as the memory resources required.

I think this could get more complicated though.

Like in cases with containers, the docker or podman daemon could be alive when glances was started but then the user shutsdown the daemon. Now glances doesnt know that it was shutdown and keeps trying to connect to it, leading to a bunch of failed connection logs.

You may think that we should stop the plugin after the daemon gets shutdown. But if the daemon starts back again, you would get not the stats in that case. So it gets very nuanced in certain situations as to what needs to be done 🤔