New Logoff function for Multi run job

[samprapist]

Configs that generated a very large amount of bot status data, such as hits and others, used a large amount of RAM unnecessarily.

The logoff option allows you to disable bot data writing, while the number of hits and other statuses is displayed.

Hits and other bot statuses can be read by saving the data to a file when creating the config.

The difference in RAM consumption for me by openbullet is:

without "logoff", 4 GB of RAM usage after a few hours

With "logoff" function enabled, 600 mb of RAM usage after a few hours.

optional: add logoff selection mode, no logs in multi run or no logs in multirun and in the database.

[openbullet]

Hello, thanks for the PR.

Unless your config is capturing a lot of data, the RAM consumption should mostly be attributed to the bot logs. Do you have Bot Log disabled in the RL Settings? Because that's probably the cause of the huge RAM spikes. Sure, it's not a per-job toggle, but if you disable the Bot Log you should still be able to save a lot of RAM, right?

Note: next time, please discuss the feature with me first, otherwise you end up writing code that may not get merged.

[samprapist]

Hello, thanks for the PR.

Unless your config is capturing a lot of data, the RAM consumption should mostly be attributed to the bot logs. Do you have Bot Log disabled in the RL Settings? Because that's probably the cause of the huge RAM spikes. Sure, it's not a per-job toggle, but if you disable the Bot Log you should still be able to save a lot of RAM, right?

Note: next time, please discuss the feature with me first, otherwise you end up writing code that may not get merged.

Yes, I had bot logs disabled in RL settings during a session in which openbullet was using 4 GB of RAM. Probably the problem with RAM consumption is only indirect with the display of bot statuses because RAM consumption increases after a long time, so the direct problem is somewhere else. Anyway, I didn't contact this suggestion before creating the code because I made it for my own needs and added it to GitHub because while reading the forum I saw that some people were interested in a similar function because they also had a problem with memory leak.