tmsh sys config save and its effect on existing production traffic

> My question is if i go in and modify the bigip.conf and then to a tmsh sys config save/load will it affect the traffic that is currently on the box.

Saving the config after hand-editing the bigip.conf will overwrite your changes, so will any changes made on the GUI. To load the specfic components of the config I would create a new file and copy/paste the items you want to edit into it, change the configuration in that new file, then load it with the "tmsh load sys config file merge" command.

The 'tmsh load sys config file ... merge' command acts very similiarly to the older 'bigpipe merge' command. Ithink think there may be a difference in the handling of duplicate objects (error vs overwrite), but I haven't used it in a while and don't remember exactly.

Saving the configuration does not have an impact on performance. The caveat might be that if you are using modules that are heavy on disk-based I/O, the time required to save the files could potentially be visible to very high resolution monitoring systems (< 10 second polling frequency), or performance test equipment. But in general the cost of saving the configuration is negligible.

Using the 'tmsh load sys config' command does have an impact on existing traffic. How much of an impact is dependent on how large the configuarion you are loading is. With small-ish configurations I can do this during a performance test and see a small, but definitely visible, impact. With large configurations the impact will come at the end of the load process when the config is pushed from mcpd to the tmm instances. It's not crippling by any means, but much more visible to high-resolution monitoring than saving. IIRC, this has no impact on the connection table and TCP should be able to recover without a problem.

I would recommend the 'tmsh load sys config file ... merge' command for the configuration changes you mention above. That command appeared in v11.1, I believe. Good luck.